How to Compose Great Prose: Cicero, Dionysius of Halicarnassus and Stylistic Theory in Late-Republican and Augustan Rome Issue Date: 2019-10-23

Home , 324 BC, Apollonius Molon, Atticism, Grand style (rhetoric), Iambus (genre), Ludwig Radermacher

Cover Page

The following handle holds various files of this Leiden University dissertation: http://hdl.handle.net/1887/79945

Author: Ooms, S Title: How to compose great prose: Cicero, Dionysius of Halicarnassus and stylistic theory in Late-Republican and Augustan Rome Issue Date: 2019-10-23

OW TO OMPOSE REAT ROSE H C G P

Cicero, Dionysius of Halicarnassus and Stylistic Theory in Late-Republican and Augustan Rome

Proefschrift

ter verkrijging van de graad van Doctor aan de Universiteit Leiden, op gezag van de Rector Magnificus prof. mr. C.J.J.M. Stolker, volgens besluit van het College voor Promoties te verdedigen op woensdag 23 oktober 2019 klokke 16.15 uur

door

Steven Ooms geboren te Breda in 1989

Promotor Prof. dr. I. Sluiter

Co-promotor Dr. C.C. de Jonge

Promotiecommissie Prof. dr. K. De Temmerman (Universiteit Gent) Dr. L. van Gils (Vrije Universiteit Amsterdam) Dr. C.H. Pieper Prof. dr. A.B. Wessels

ISBN 978-94-6375-519-1. Printed by Ridderprint BV. Cover design by Elisa Calamita.

This dissertation is part of the research project ‘Greek Criticism and Latin Literature: Classicism and Cultural Interaction in Late-Republican and Early-Imperial Rome’, funded through a Vidi grant by the Dutch Research Council (NWO), project number 276-30-009.

voor

MATHILDE

Contents

Preface ix Note on translations and conventions xi

1. General introduction: Cicero and Dionysius of Halicarnassus on prose style 1

1.1 The purpose of this study 1 1.2 Modern approaches to the stylistic theories of Cicero and Dionysius 4 1.3 Theories of style in Rome: four major themes 11 1.4 Latin stylistic theory: Cicero, Calvus and Rhetorica ad Herennium 13 1.5 Greek stylistic theory in Rome: Dionysius, Caecilius and Philodemus 20 1.6 Athens into Rome: the shared classicism of Cicero and Dionysius 29 1.7 The dialogue between Greek and Latin stylistic theory 36

2. Hero versus zero: Demosthenes and Hegesias as paradigms of good and bad style 43

2.1 Introduction 43 2.2 Demosthenes and Atticism, Hegesias and Asianism 46 2.3 How to compose great prose: Demosthenes as the canon of style 52

2.3.1 Demosthenes in Hellenistic scholarship 53 2.3.2 Cicero on Demosthenes’ style 56 2.3.3 Dionysius on Demosthenes’ style 61 2.3.4 Summary: the three principal stylistic virtues of Demosthenes 69

CONTENTS

2.4 How not to compose great prose: Hegesias as the prototype of degeneracy 70

2.4.1 Agatharchides of Cnidus on Hegesias’ style 71 2.4.2 Cicero on Hegesias’ style 78 2.4.3 Dionysius on Hegesias’ style 86 2.4.4 Summary: the three principal stylistic vices of Hegesias 94

2.5 Conclusion 95

3. The plain, the grand and the in-between: The doctrine of three styles as a versatile critical tool 97

3.1 Introduction 97 3.2 Versatile systems of three styles 104

3.2.1 Malleable dichotomies: the stylistic extremes 106 3.2.2 The flexibility of the middle style 112

3.3 Classicism and the three styles 115 3.4 Cicero against the so-called Attici 122 3.5 Dionysius and the aesthetics of the middle 129 3.6 Greek and Roman audiences: the three styles in action 139 3.7 Conclusion 145

4. Offending the ears: Greek and Latin views on rough word arrangement 147

4.1 Introduction 147 4.2 Word arrangement: a Greek discipline in Rome 153 4.3 The universal law of arrangement and the judgment of the ear 161 4.4 Disrupting the natural order: sources for roughness 168

CONTENTS

4.5 Echoing greatness: the virtues of roughness in Greek and Latin prose 175

4.5.1 Simplicity and sincerity 176 4.5.2 Grandeur and sublimity 181 4.5.3 Archaism and the patina of antiquity 186

4.6 Conclusion 192

5. To be Attic or not to be Attic: The fluidity of Atticism in oratory, politics and life 195

5.1 Introduction 195 5.2 Atticism in modern scholarship 197 5.3 Attic style and the Athenian cultural repertoire 203 5.4 Atticism personified: the various Attic muses of Rome 211 5.5 The politics of stylistic theory: Cicero and Dionysius on Attic style 216

5.5.1 Cicero’s sublime oratory and the battle for the republic 217 5.5.2 Dionysius’ middle style and the rule of Augustus 223

5.6 Calvus’ Atticism: challenging Rome’s rhetorical and political elites 231

5.6.1 Calvus against the rhetorical status quo 232 5.6.2 Calvus against public and private effeminacy 242

5.7 Conclusion 246

6. General conclusion 249

Bibliography 255 Samenvatting (in Dutch) 285 Curriculum vitae (in Dutch) 294

vii

Preface

The power of words can hardly be overstated: who among us has not been inspired or devastated, excited or consoled, moved to tears or otherwise touched by a felicitous phrase? It is no wonder, then, that verbal proficiency is a much sought-after commodity today: novelists strive to write sentences that take your breath away, politicians hunt for one-liners to knock out their opponents, and internet users dream of tweeting a post that conquers the online community. Although the quest for great prose is highly topical, I propose that it should start in Antiquity. Even after thousands of years, the Greek and Latin classics do not cease to engross billions of readers worldwide: by the sheer magnitude and longevity of their success, these age-old masterpieces offer essential clues to attaining enduring literary fame. Therefore, I do not regret spending more than five years on two of the most renowned ancient experts on verbal excellence, the Roman superstar Cicero and the versatile Greek scholar Dionysius of Halicarnassus. This dissertation reports what they have to say about the issue that continues to fascinate me—how to compose great prose. My research was funded through a Vidi grant by the Dutch Research Council (NWO). Since I started working on this project in February 2014, I have benefited from the convivial and stimulating atmosphere of the Leiden University Centre for the Arts in Society (LUCAS): I am referring particularly to the companionship and talent of my fellow PhD students, and to the kindness and expertise of my colleagues in the Classics Department. Professor Ineke Sluiter and dr. Casper de Jonge have on many occasions read and corrected my less-than-great prose: their prudent counsel and unwavering support have kept me on track despite my tendency to get derailed. I am also indebted to the National Research School in Classical Studies (OIKOS) for allowing me to pitch my preliminary findings to various (inter)national audiences in the Netherlands and in Athens, at the very foot of the Acropolis. In addition, I warmly thank my colleagues at the Erasmiaans Gymnasium in Rotterdam, who offered me a genial home for the final stages of my research. As for my students, I can only hope that I will some day be able to repay them for everything that they have taught me. Writing a doctoral thesis comes with its own trials and triumphs: I count myself lucky that I faced the former and celebrated the latter together with my friends and family. I am deeply grateful to Marianne Schippers, who has constantly cheered me up and spurred me on, and to my amigos Shane Kolenbrander and Vincent Oort, whose laconian wit has helped me in many ways. I also wholeheartedly thank my stepfather Willem van Heusden, my brothers

PREFACE

Mark and Jeroen, my grandfather Piet and my late grandmother Corrie for their comforting presence and patience. Lastly, I owe an enormous debt of gratitude to my mother and my wife. My mother Wilma Lambregts has been my rock for over thirty years: if I have achieved anything at all, it is because of the loving confidence that she has always placed in me. My wife Jennifer Snellink is both my better half and my best friend: God only knows what I’d be without her. She corrected countless mistakes in an earlier draft of this dissertation; the remaining errors (both in print and in life) are all mine. It is my contention that the importance of well-chosen and well-ordered words is timeless: eloquence was a paramount issue in Antiquity, it still is in our day and it will remain so in the future. On that note, I think of my daughter Mathilde, who was born in January this year: up to now she has only emitted inarticulate gibberish, but soon she will be forming actual words, on her way to compose her own great prose. In accordance with the ancient wisdom of Cicero and Dionysius, I hope that she will find a voice that highlights her virtues, that suits her talents, and that helps her achieve all of her ambitions, whatever they may be. Therefore, sweet, babbling Mathilde, I dedicate this book to you.

Voorschoten Summer 2019

Note on Translations and Conventions

1. When translators are not mentioned by name, translations are drawn (and often adapted) from the Loeb series. For Agatharchides’ On the Red Sea , I have used Burstein (1989).

2. Abbreviations for modern works of reference as well as ancient authors and their works are generally borrowed from The Oxford Classical Dictionary (OCD , fourth edition), and otherwise from A Greek-English Lexicon (LSJ, ninth revised edition) or The Oxford Latin Dictionary (OLD , second edition).

3. For Dionysius’ works, the following abbreviations are used:

Latin English Amm. I Epistula ad Ammaeum I First Letter to Ammaeus Amm. II Epistula ad Ammaeum II Second Letter to Ammaeus Ant. orat. De antiquis oratoribus 1 On the Ancient Orators Ant. Rom. Antiquitates Romanae Roman Antiquities Comp. De compositione verborum On the Arrangement of Words 2 Dem. De Demosthene On Demosthenes Din. De Dinarcho On Dinarchus Imit. De imitatione 3 On Imitation Is. De Isaeo On Isaeus Isoc. De Isocrate On Isocrates Lys. De Lysia On Lysias Pomp. Epistula ad Pompeium Letter to Pompeius Thuc. De Thucydide On Thucydides

1 Also known as De oratoribus veteribus (Orat. vett. ). 2 Often referred to as On Composition . 3 Sometimes listed as De veterum censura (Vett. cens. ).

4. References to Dionysius’ rhetorical-critical treatises follow the numbers of chapters and sections in the Budé editions of Aujac (1978–1992).

Chapter 1 GENERAL INTRODUCTION : CICERO AND DIONYSIUS OF HALICARNASSUS ON PROSE STYLE

1.1 The Purpose of this Study At first sight, M. Tullius Cicero (106–43 BC) and Dionysius of Halicarnassus (ca. 60–after 8 BC) 1 have little in common. Both men made their career in Rome during the first century BC, but their paths never actually crossed: when Dionysius settled in the city around 30 BC, Cicero had been dead for well over a decade. 2 When Cicero composed his last speeches, letters and treatises, Rome was still a republic, but when Dionysius started to teach rhetoric in Rome, the city was ruled by Octavian, who emerged as emperor under the title Augustus. In addition, Cicero is considered an exponent of Roman eloquence and a master of Latin prose, whereas Dionysius was a Greek historian and a critic of Greek literature. As a consequence of these apparent dissimilarities, modern scholars, with a few notable exceptions (section 1.2), rarely mention Cicero and Dionysius together in the same paragraph, section or chapter. Still, their names are printed jointly on the cover of the present book, which is concerned with stylistic theory. Indeed, it is on the topic of style, particularly prose style, that the views of Cicero and Dionysius lend themselves to a careful comparison. In ancient rhetorical theory, style ( λέξις , elocutio ) was usually considered the third task of the orator, after invention ( εὕρεσις , inventio ) and arrangement ( τάξις , dispositio ) of the subject matter: while the two prior tasks focus on finding and organizing one’s materials and one’s thoughts, style is all about putting these into words. In Late-Republican and Augustan Rome, Cicero and Dionysius stand out as our most important sources for stylistic theory: their

1 Dionysius himself claims that Plato (born in 427 BC) was his senior by more than twelve generations (Pomp. 1.15) and he tells us that Crassus’ expedition against the Parthians (55–53 BC) took place in his time ( Ant. Rom. 2.6.4). Most scholars assume that Dionysius was born around or shortly after 60 BC: see e.g. Hidber (1996) 2 and Fromentin (1998) xiii. The last known date from Dionysius’ life is 8 or 7 BC (the consulship of Claudius Nero and Calpurnius Piso), when he published the first book of Ant . Rom . (cf. ibid. 1.3.4): see Cary (1937) vii and Hidber (1996) 1. Dionysius probably lived on for quite some time after 8/7 BC, as he had yet to publish the remaining nineteen books of Ant. Rom . 2 Dion. Hal. Ant. Rom. 1.7.2: ‘I arrived in Italy at the very time when Augustus Caesar put an end to the civil war, in the middle of the one hundred and eighty-seventh Olympiad ...’ ( ἐγὼ καταπλεύσας εἰς Ἰταλίαν ἅμα τῷ καταλυθῆναι τὸν ἐμφύλιον πόλεμον ὑπὸ τοῦ Σεβαστοῦ Καίσαρος ἑβδόμης καὶ ὀγδοηκοστῆς καὶ ἑκατοστῆς ὀλυμπιάδος μεσούσης ...). Thus, Dionysius arrived in Italy in 30 or 29 BC: see Hidber (1996) 1–2.

CHAPTER ONE rhetorical and critical treatises are virtually the only surviving works from that era that include lengthy discussions of the intricacies of prose style (sections 1.4 and 1.5). Moreover, their interest in style was by no means purely academic, as both men contributed to the city’s prolific output of artistic prose: Cicero’s works, for one, quickly became stylistic models in their own right for generations to come, while Dionysius’ literary aspirations were enshrined in his twenty-volume history of early Rome, commonly known as his Roman Antiquities (Ῥωμαϊκὴ Ἀρχαιολογία ). Cicero and Dionysius offer us unique first-hand perspectives on prose style in the last fifty-or-so years BC, a time traditionally celebrated as the Golden Age of Latin literature, 3 during which Greek literature (according to Dionysius) also flourished.4 The rhetorical and critical works of Cicero and Dionysius can be read as masterclasses on how to compose great prose in the literary heyday of Late-Republican and Augustan Rome. By comparing their stylistic views, this study aims to reconstruct the ancient discourse on prose style in Rome. Specifically, I will focus on the four major themes that permeate the stylistic discussions in the works of both Cicero and Dionysius, namely the selection of models for imitation (chapter 2), the use of threefold stylistic divisions (chapter 3), theories of word arrangement (chapter 4) and the ideology of Atticism (chapter 5). Each of these subsequent chapters substantiates the three central theses of this dissertation:

1. By and large, Cicero and Dionysius draw on similar stylistic theories, they apply similar analytical methods and they articulate their views using a similar technical vocabulary. Elements from this shared apparatus can be found in various other contemporary or near-contemporary discussions of style in Rome, such as Rhetorica ad Herennium and the sources on the orator C. Licinius Macer Calvus (section 1.4) as well as, in Greek, the fragments of Caecilius of Caleacte and Philodemus’ On Rhetoric and On Poems (section 1.5). In sum, the various extant stylistic discussions in Rome draw on and contribute to a common discourse on prose style.

3 The Golden Age of Latin literature is usually defined as the period between the start of Cicero’s career (ca. 80 BC) and the end of emperor Augustus’ reign (14 AD), followed by the so-called Silver Age. This classification, based on the ancient myth of the ages of man, was first articulated in the nineteenth century by Teuffel (1870) and Cruttwell (1877). Although the value judgments, which are implied in the words ‘golden’ and ‘silver’, are controversial, the terms are still used in modern scholarship, e.g. Wilkinson (1963): Golden Latin Artistry. 4 At Ant. orat. 3.2, Dionysius notes that in his day many fine works ( συντάξεις ) of literary prose ‘have proceeded from the pens of Romans and Greeks and will probably continue to do so’ ( καὶ Ῥωμαίοις καὶ Ἕλλησιν εὖ μάλα διεσπουδασμέναι προεληλύθασί τε καὶ προελέυσονται κατὰ τὸ εἰκός ).

GENERAL INTRODUCTION

2. The shared framework of stylistic theories, methods and vocabulary was by no means rigid or monolithic, but rather flexible and fluid. Hence, Cicero, Dionysius and their contemporaries could articulate diverse ideas about style, whilst using the same basic critical tools: they could bend the commonplace topics of the stylistic discourse to suit their personal aesthetic tastes, to pursue specific writing purposes, to cater for their intended audiences, to reflect differences between the Greek and Latin language, and, lastly, to demarcate their own position as authors and citizens in Rome.

3. The common flexible language of style allowed Greek and Roman authors alike to participate in a mutual exchange of ideas about the topic. 5 In this dissertation, we will see Greeks and Romans ‘interacting rather than assimilating’: 6 the apparent connections between Greek and Latin sources should not be taken as evidence for the hellenization of Roman rhetoric or for the merging together of Greek and Latin stylistic theory. Rather, we will see that both Greek and Roman authors exploited a common cultural repertoire, or ‘koine’ (to borrow a useful concept from recent scholarship on material culture) in order to communicate views about the present-day Roman society by focusing on the rich legacy of the Greek literary past (section 1.7).7

In a word, Cicero and Dionysius participated in and contributed to a flexible discourse on prose style, shared by Greeks and Romans alike, through which they not merely exchanged ideas about style and literature, but also negotiated their identity in the public life of Rome. In the remainder of this introductory chapter, I will explain how my approach to Cicero and

5 I am aware of the inherent ambiguity of the terms ‘Roman’ and ‘Greek’. Scholars have emphasized the fluidity of these categories, distinguishing between ethnic and cultural identities: it is possible, for instance, that ancient authors present themselves as Greeks or Romans, while they are in fact not originally from Greece or Rome. See e.g. the seminal work of Hall (2002) on the construction of Greek identity during the Persian Wars, and Dench (2017) on the interplay of ethnicity, culture and identity among Greeks living under Roman rule during the Second Sophistic. In this dissertation, I simply refer to the authors of Latin texts as ‘Roman authors’ and to the authors of Greek texts as ‘Greek authors’: as we will see (cf. n. 141 below), these authors often explicitly present themselves as Romans (Cicero, and the author of Rhet. Her. ) or as Greeks (Dionysius, Longinus). 6 Cf. Wallace-Hadrill (1998) 84. 7 Wallace-Hadrill (2008) 3–37 offers a thorough discussion of the historical debates over the interrelations between hellenization and romanization: he debunks popular models such as fusion, creolization, hybridity and other metaphorical representations that rely on the model of acculturation. On the concept of a cultural ‘koine’, see section 1.7 below and the literature cited there.

CHAPTER ONE

Dionysius builds on and reacts to existing scholarship on ancient stylistic theory (section 1.2), I will offer an introduction to the relevant texts about prose style from Late-Republican and Augustan Rome (sections 1.3–1.5) and I will discuss the classicism in the works of Cicero and Dionysius, providing a convenient starting point for exploring the dialogue between Greek and Roman authors in the ensuing chapters (sections 1.6–1.7).

1.2 Modern Approaches to the Stylistic Views of Cicero and Dionysius This dissertation is not the first study to point out the remarkable connections between the stylistic views of Cicero and Dionysius of Halicarnassus. As early as 1910, the German classicist Franz Nassal published his dissertation on the ‘aesthetic and rhetorical connections’ (‘aesthetisch-rhetorische Beziehungen’) between the two authors, focusing particularly on their kindred views on style. 8 Nassal’s monograph deserves to be mentioned here for two reasons: first, it is the only work to date that presents a detailed and systematic comparison between the stylistic views of Cicero and Dionysius, and secondly, it is the first study to explicitly offer an explanation for the striking parallels between them. 9 Nassal argues that Cicero and Dionysius must rely on a common source, that can only be ‘a work by a highly educated Greek’. 10 In Nassal’s assessment, this Greek ancestor was probably Caecilius of Caleacte, whose rhetorical and critical works are only known to us through testimonies and a handful of fragments (section 1.5). In all likelihood, however, Caecilius was younger than Cicero: the Byzantine encyclopedia Suda states that he flourished in Augustan Rome and Dionysius calls him a ‘very dear friend’. 11 Therefore, Nassal’s attempt to make him Cicero’s and Dionysius’ senior as well as their common source is unconvincing. 12

8 Nassal (1910) . The book is divided into two main sections, the first focusing on Cicero’s and Dionysius’ ‘rhetorical and technical views’ (‘rhetorisch-technische Anschauungen’), the second on their ‘aesthetic and rhetorical judgments of Greek authors’ (‘ästhetisch-rhetorische Urteile über griechische Schriftsteller’). In both of these parts, Nassal focuses on their views on style. 9 Before Nassal, Ammon (1889) had included Cicero in his discussion of sources for Dionysius’ rhetorical treatises, but he did not do so ‘ex professo und bis in die Einzelheiten’, according to Nassal (1910) 3. 10 Nassal (1910) 6–7: ‘Ein Werk eines hochgebildeten Griechen.’ 11 Suda κ 1165 (= Caecilius T1 Woerther) states that Caecilius was a ‘rhetorician who gave lectures in Rome under Augustus Caesar and until Hadrian’ ( ῥήτωρ , σοφιστεύσας ἐν Ῥώμῃ ἐπὶ τοῦ Σεβαστοῦ Καίσαρος καὶ ἕως Ἀδριανοῦ). The reference to Hadrian cannot be correct. See also Dion. Hal. Pomp. 3.20: τῷ φιλτάτῳ Καικιλίῳ. On Caecilius’ role in the stylistic discourse in Rome, see section 1.5 below. 12 Nassal follows the argumentation of Wilamowitz (1900) 15: according to the latter, the titles of Caecilius’ works Against the Phrygians and How the Attic and Asian Styles Differ suggest that the battle between Atticists

GENERAL INTRODUCTION

Nassal’s study exemplifies two implicit presuppositions that have, for a long time, set the parameters of modern scholarship on ancient rhetoric and style. First, scholars have in the past generally taken a positivist approach to ancient stylistic theory: they perceived it as a fixed system of abstract guidelines that can be reconstructed by piecing together the evidence from the extant sources. 13 From this perspective, classicists such as Nassal have scoured the works of Cicero and Dionysius for nuggets of otherwise lost theoretical wisdom, treating Cicero and Dionysius themselves as passive recipients of earlier scholarship. 14 Secondly, in their search for the Greek foundations of stylistic theory, scholars have traditionally adopted a hellenocentric view, interpreting Latin texts as mere echoes of Greek theory. The evidence for this view, as we will see, is usually quite flimsy, as the alleged Greek sources are often not extant. 15 This problem is starkly visible in the following words of Richard Janko: ‘Cicero’s whole account of euphony and rhythm is indebted to the κριτικοί ; one could easily turn his words back into Greek.’ Flamboyantly, Janko actually produces his own Greek translation of Cicero’s Latin, thereby inventing his own proof for the claim that Cicero copied from his Greek predecessors. 16 and Asianists was ongoing at the time of publication, whereas Dionysius proclaims the victory of Atticism in Ant. orat. 1.1–7. Accordingly, Wilamowitz and Nassal reason that Caecilius’ works must predate Dionysius. Yet, as De Jonge (2008) 216 n. 205 notes, ‘we should avoid presenting the conflict between Atticists who objected to “Asianic” style as a real “battle” that was decided at a particular moment’. On the date of Caecilius, see also Bowersock (1965) 124 and Hidber (1996) 41 n. 184. On Atticism and Asianism, see esp. section 5.2 below. 13 On the positivist view of rhetorical theory, see Copeland (1991) 4: rhetoric ‘has been viewed as a neutral perceptive system, a descriptive taxonomy of style, or as an academic discipline whose history is constituted by its manifest meanings and whose claims to truth about the nature of discourse and language are accepted on their own terms.’ Lausberg (2008) is a good example of this approach, as he aims to provide a comprehensive description of the ancient rhetorical system by collecting material from a wide variety of sources. While the book is an extremely useful reference work, it tends to elide important differences between the ancient sources, giving the false impression that ancient rhetoric functioned as a unified system. 14 See May and Wisse (2001) 38–39 on the limitations of source criticism in the study of Cic. De or. See also De Jonge (2008) 7–9 on the same issue in the study of Dionysius’ rhetorical works. 15 We will encounter instances of the hellenocentric approach on several occasions in this dissertation: see e.g. section 3.1 (on the three styles), section 4.1 (on the theory of word arrangement) and section 5.2 (on Atticism). 16 Janko (2000) 361 n. 3. The passage is Cic. Orat. 162: Sed quia rerum verborumque iudicium in prudentia est, vocum autem et numerorum aures sunt iudices, et quod illa ad intelligentiam referuntur, haec ad voluptatem, in illis ratio invenit, in his sensus artem. Janko’s translation reads: Ἐπεὶ δ’ ἡ μὲν τῶν πραγμάτων καὶ τῶν ὀνομάτων κρίσις ὑπάρχει ἐν τῇ νοήσει , τὰς δὲ φωνὰς καὶ τοὺς ῥυθμοὺς κρίνουσιν αἱ ἀκοαί , καὶ ἐπεὶ τὰ μὲν ἐπὶ τὴν διάνοιαν ἀναπέμπεται , τὰ δὲ πρὸς τὴν τέρψιν , ἐν ἐκείνοις μὲν ἡ διάνοια ἐξεῦρε τὴν τέχνην , ἐν τούτοις δ’ ἡ αἴσθησις . On the views of οἱ κριτικοί , who focused on euphony and word arrangement, see esp. section 4.1 below.

CHAPTER ONE

To be fair, the scholarly trends signaled above are by no means pertinent to every single study of ancient stylistic theory from the nineteenth and twentieth centuries: I have simply aimed to describe general tendencies. 17 At any rate, toward the end of the millennium the predominant scholarly approach started to shift: this change of attitude can be illustrated in the following selection of recent work on Ciceronian and Dionysian rhetoric. 18 Among the students of Cicero, to begin with, the hellenocentric bias has in recent decennia lost ground: in the first-ever English monograph on Cicero’s De Oratore , for one, Elaine Fantham emphatically situates Cicero’s theories in the ‘Roman world’. 19 Likewise, the positivist focus on source criticism is now largely abandoned, clearing the path to study the goals and motivations that underlie Cicero’s rhetorical treatises: following Emanuele Narducci’s suggestion that these works are engaged in a conscious cultural program, 20 John Dugan interprets Cicero’s views in the light of his program of self-fashioning, making his rhetorical theory an expression of his identity as a ‘new man’. 21 Others, in addition, have highlighted the political dimensions of Ciceronian rhetoric: Joy Connolly discusses Cicero’s views with respect to their civic and political context, 22 while Caroline Bishop, whose dissertation

17 A case in point is the longstanding debate over the origins of Atticism: although most scholars of the nineteenth and twentieth centuries have focused on the allegedly Greek roots of the phenomenon, various classicists have, with varying degrees of confidence, suggested that it might have originated as a Roman movement. See Radermacher (1899a) 360, Heck (1917), Kennedy (1972) 245–246 and 351–353, Bowersock (1979) 67, Innes (1989) 245–246 and Wisse (1995). For their views, see section 5.2 n. 21 below. 18 For general trends in the scholarship on Ciceronian rhetoric, see Dugan (2007) and Dugan (2013) 25–27. An up-to-date overview of literature on Dionysius’ critical works is offered by De Jonge and Hunter (2019) 17–24. 19 Fantham (2004) studies the ‘Roman world of Cicero’s De Oratore ’. Fantham aims to illuminate the political climate in Rome in the period between the dialogue’s fictional date (91 BC) and its date of composition (55 BC). In chapter 7, Fantham argues that Cicero reviewed Aristotelian rhetorical theory in the light of the judicial and political situation in Rome. 20 Narducci (1997) collects five articles on Cicero’s rhetorical theory that aim to illuminate the ideological presuppositions that control Cicero’s construction of the orator and his culture. Narducci’s discussions are more suggestive than conclusive: cf. the reviews in Dugan (1998) and Steel (1998). 21 Dugan (2005) reads De or., Brut. and Orat. not only as sources for Ciceronian rhetorical theory but also as sources for the author’s ‘cultural program’. According to Dugan, Cicero cultivates a ‘transgressive’ approach to aesthetics: in his view, Cicero presented himself as a ‘new man’ by subverting established aesthetic tastes and oratorical customs, e.g. by adopting a stagey, highly emotional instead of a subdued style. See however the review of Connolly (2006) on the limited usefulness of the notion of transgression in Cicero’s works. 22 Connolly (2007a) analyzes Latin rhetorical works, especially Cicero’s, as serious analyses of the role of the citizen and as practical guides for their readers toward a contructive integration of rhetoric with public life. Here and more fully in Connolly (2007b), the author also addresses the obsession of ancient orators and rhetoricians

GENERAL INTRODUCTION examines Cicero’s engagement with Greek scholarship, understands the rhetorical treatises Brutus and Orator as reactions to Caesar’s dictatorship. 23 In recent scholarship on Dionysius, similar trends can be detected. The Greek critic’s works are no longer studied exclusively in connection to the Greek tradition: according to Jakob Wisse, for example, Dionysius’ views on Attic style were influenced by Roman ideas that reached him through a ‘Graeco-Roman network’ of scholars in Rome. 24 In addition, scholars have gradually become more willing to contemplate the originality and individuality of Dionysius’ views: Koen Goudriaan has argued, for instance, that a personal program of civilization underlies Dionysius’ historical and critical works, 25 and Nicolas Wiater has more recently interpreted Dionysius’ entire oeuvre as a model for Greek cultural identity, through which the Greek scholar presents the Romans as dependent on the superior culture of Classical Greece. 26 Additionally, Casper de Jonge has shown that the various theories in Dionysius’ critical treatises do not merely reiterate earlier views, but rather that Dionysius selects theories for a practical reason, that is, to instruct his audiences on the composition of effective texts. 27 Dionysius’ works are now often reviewed in the context of Augustan Rome: a brand-new volume explores the complex ways in which Dionysius’ oeuvre fits into the

with gender: they often presented themselves as thoroughly virile and their opponents as emasculated or effeminate. Cf. on this latter point sections 5.6.1 and 5.6.2 below. 23 Bishop (2011) explores Cicero’s emulation of the Greek authors Aratus, Plato and Demosthenes. Bishop (2016) shows that Cicero presents himself as a Roman Demosthenes in Brut. and Orat. : just as Demosthenes vociferously attacked Philip of Macedon, Cicero in his day struggled with the rule of Caesar. In his Phil. of 44– 43 BC, Cicero famously uses the eponymous speeches of Demosthenes as his model: see esp. Wooten (1983). 24 Wisse (1995) distinguishes between a Roman phase of Atticism (Calvus and Brutus) and a subsequent Greek phase (Dionysius and Caecilius): according to Wisse, ‘it was in the coterie of Calvus that Atticism took shape ca. 60 BCE; it spread through a Graeco-Roman network, to emerge, for us, as a Greek phenomenon in Dionysius of Halicarnassus’ (ibid. 81). For the debate about the origins and nature of Atticism, see esp. section 5.2 below. 25 Goudriaan (1989) explores Dionysius’ classicism in his critical treatises as well as Ant. Rom. : according to Goudriaan, the overarching ideal of civilization ( παιδεία ) that emerges from Dionysius’ oeuvre resembles Nietzsche’s notion of ‘formale Bildung’, that is, the comprehensive education of a human being through intensive linguistic training. See esp. ibid. 581–586. 26 Like Goudriaan, Wiater (2011) focuses on the ideology of classicism: he argues that Dionysius’ works create ‘a distance between Greek and Roman readers’ (ibid. 222), supplying the former with a means to assert their superiority over the latter. Yet, the review of De Jonge (2012b) shows that there are many striking links between Dionysius’ classicist program and Roman forms of classicism: cf. sections 1.6 and 5.5.2 below. 27 De Jonge (2008) explores Dionysius’ views on language, linguistics and literature, showing that the critic combines various ancient language disciplines into a coherent program of rhetorical education.

CHAPTER ONE intellectual and political world of Rome under Augustus. 28 Among the more intriguing issues that have come to the fore are the striking connections between Dionysius’ critical doctrines and contemporary Latin poetry: following the suggestions of Richard Hunter and others, De Jonge exposes several conspicuous links between Dionysius’ treatise On the Arrangement of Words on the one hand and Virgil’s Aeneid and Horace’s Ars Poetica on the other hand. 29 Thus, the positivist, hellenocentric approach to ancient stylistic theory is largely debunked, yielding valuable new insights into the rhetorical and critical views of Cicero and Dionysius. Now, more than a century after Nassal’s dissertation, it is time for a thorough reassessment of the connections between the two ancient authors: how can we explain the striking similarities between their views about prose style? Whereas it is chronologically impossible for Cicero to have known Dionysius’ works, some scholars have suggested that Dionysius drew on the works of Cicero. 30 Glen Bowersock, for instance, has argued that Dionysius must have been aware of Cicero’s ideas through Q. Aelius Tubero, who was the recipient of Dionysius’ treatise On Thucydides and who also knew Cicero through his father Lucius. 31 Indeed, even without Tubero, it seems hardly possible for someone like Dionysius, who was actively involved in Roman society, not to be au fait with Cicero’s ubiquitous legacy. Therefore, De Jonge lists Cicero as a probable source for Dionysius’ works: ‘Despite the modern reluctance to make a Greek scholar dependent on a Roman author, parallels

28 De Jonge and Hunter (2019): the editors present Dionysius as ‘a writer positioned between Greece and Rome and between rhetoric and historiography’. De Jonge and Hunter divide the topics of the various contributions to the volume into four categories, i.e., Augustan politics, Augustan historiography, Augustan rhetoric and, lastly, Augustan criticism and Latin poetry. 29 De Jonge (2018a) calls attention to the ideal of noble simplicity, which plays a role in Comp. (through the swineherd Eumaeus), in Verg. Aen. (through king Evander) and in Augustus’ self-presentation as a good ruler. De Jonge (2019) shows that both Dion. Hal. Comp. and Hor. Ars. P. advocate the skillful arrangement of common words: cf. section 4.1 n. 4 below. Freudenburg (1993) 109–184, Fuhrer (2003) 355–356, and Hunter (2009) 126–167 already suggested that Dionysius’ theory of word arrangement can shed new light on Hor. Sat . According to Hunter, Dionysius and Horace ‘are in touch with similar streams of criticism’ (ibid. 165). 30 Cf. already Ammon (1889), who included Cicero in his list of Dionysius’ sources, and Egger (1902) 77, who briefly considered the possibility that Dionysius drew on Cicero for his theory of word arrangement. 31 Bowersock (1965) 130 and Bowersock (1979) 68–70. Cicero refers to his relationship with both Lucius and Quintus throughout his Pro Ligario . At Lig. 1, he calls Quintus ‘my kinsman’ ( propinquus meus ) and at Lig. 12 and 21, he refers to his friendship with Lucius and Quintus. Dionysius mentions Quintus at Thuc. 1.1, Amm. II 1.1 and Ant. Rom. 1.80.1. There is no evidence for Bowersock’s claim that Q. Aelius Tubero was Dionysius’ patron: see De Jonge (2008) 28. Wisse (1995) 78 rightly notes that we should not suppose that Dionysius’ knowledge of Roman ideas was dependent on any single individual like Tubero.

GENERAL INTRODUCTION between Cicero and Dionysius may be based not only on their use of earlier theories, but also on Dionysius’ knowledge of Cicero’s treatises.’ 32 Still, it is difficult to find ironclad evidence for direct Ciceronian influence in Dionysius’ works: the Greek critic never explicitly refers to Cicero and, whenever he expounds stylistic doctrines that Cicero also mentions, these doctrines are, as we will see, usually attested in other contemporary sources as well. To support the argument that Dionysius knew Cicero’s works, a passage from On Thucydides is sometimes adduced, in which Dionysius rejects the view of ‘some professors of repute’ ( τινες οὐκ ἄδοξοι σοφισταί ) that the work of Thucydides may serve as a good model for the writing of history. 33 This contested view corresponds with the one that Cicero presents in Brutus and Orator : according to Cicero, Thucydides is a paragon of serious and dignified historiography, though his style is unsuitable for oratory. 34 It is not unlikely that Dionysius is thinking of Cicero as one of the ‘professors of repute’, but Thucydides was a much-discussed author in Rome and the various arguments about his suitability as a model were widely available. 35 A passage from the preface to On the Ancient Orators presents a similar case: Dionysius describes Asian rhetoric as a ‘Mysian or Phrygian or a Carian creature’, seemingly echoing Cicero’s reference to the same three regions in his critique of Asianism.36 Again, Dionysius may allude to Cicero, but it is equally possible that he simply repeats a cliché of their shared stylistic discourse. 37

32 De Jonge (2008) 41. 33 Dion. Hal. Thuc. 50.2–3. 34 Cic. Brut. 287, Orat. 31. Goudriaan (1989) 266 thinks that Cicero’s praise of Thucydidean historiography is insincere (‘onoprecht’), because he elsewhere prefers Isocrates and Theopompus as stylistic models ( Orat. 207). Yet, De Jonge (2008) 215 n. 199 rightly objects that Cicero’s preference for the periodic style of Isocrates and Theopompus does not preclude the use of Thucydides as a model, whom he praises for different virtues. 35 The identity of the ‘professors of repute’ is a topic of much debate: Cicero, Didymus and Tubero have been suggested as possible referents. See for a recent survey of the discussion De Jonge (2008) 215–220, who argues that ‘we should at least mention the possibility that Cicero was one of them’. Porter (2016) 217 shows that in the first century BC the work of Thucydides attracted ‘an entire critical industry’, whose origins are unknown but whose terminology and thinking ‘was completely available to Aristotle and Theophrastus, and even earlier’. De Jonge (2017) connects Dionysius’ attitude to Thucydides with the latter’s reputation in Rome. 36 Dion. Hal. Ant. orat. 1.7: Μυσὴ ἢ Φρυγία τις ἢ Καρικόν τι κακόν . Cf. Cic. Orat. 25: Caria et Phrygia et Mysia. Bowersock (1979) 65–66 notes that Dionysius repeats Cicero’s ‘refrain’, while De Jonge (2008) 15 suggests that ‘it is not impossible that Dionysius knew Cicero’s ideas and alludes to them’. 37 Hidber (1996) 111 and De Jonge (2008) 15 rightly explain that Caria, Phrygia and Mysia stand for Asian style in general. Cf. Cic. Orat. 57 ( Phrygia et Caria ). The title of Caecilius’ work Against the Phrygians underscores the status of Phrygia as a shorthand for Asian rhetoric: see section 1.5 below. Cf. also Cic. Flac. 65.

CHAPTER ONE

Hence, although it is perfectly possible, even plausible, that Dionysius read and used Cicero’s rhetorical works, it is impossible to prove it. It would be better, therefore, not to focus on ambiguous vestiges of direct influence, but rather to assume that Cicero, Dionysius and their colleagues tapped into a shared repertoire of stylistic theories, techniques and terminology, through which they could articulate and exchange ideas about prose style. To summarize, this dissertation is built on the following basic principles of method that correlate with each of the three central theses listed above (section 1.1):

1. This study is no ‘Quellenforschung’: although source criticism can be extremely useful in revealing the origins and the historical development of ancient doctrines, this dissertation aims to reconstruct the Greek and Roman stylistic discourse in a specific time at a specific place. Therefore, (a) I will adopt a synchronic approach to ancient stylistic theory, focusing on Late- Republican and Augustan Rome (roughly first century BC); (b) I will not attempt to map the direct lines of influence between individual authors in Rome, as such connections can seldom be demonstrated unequivocally. Instead, the various parallels between Cicero, Dionysius and others will be treated as elements of a common critical discourse on prose style.

2. The elements of this shared stylistic repertoire will not be read as the monolithic dogmas of any definitive theory. Unlike traditional positivist readings, therefore, this study does not understand ancient stylistic theory as a neutral system of fixed doctrines, but rather as a construct that ancient rhetoricians and critics could adopt and adapt to suit their own preferences, purposes and programs. 38

3. In exploring the intriguing cross-connections between Greek and Roman views on prose style, this study presupposes no intrinsic hierarchy between Greek and Latin texts. All sources, whether they be written in Greek or in Latin, should each be discussed as contributions to Rome’s stylistic discourse in their own right. From this vantage point, we will see that Greek and Roman scholars participated in a dialogue, each negotiating their own position in the city as authors and as citizens on the basis of a shared repertoire of theories, techniques and terminology.

38 Cf. Yunis (2019) 86–87: ‘Although the common ground between Cicero and Dionysius on this issue (i.e., Atticism) is considerable, it is limited in ways that reflect the different contexts in which they operated.’

GENERAL INTRODUCTION

1.3 Theories of Style in Rome: Four Major Themes In the previous sections, I have referred several times to the shared discourse in which the Greek and Latin discussions of prose style in Late-Republican and Augustan Rome participate. Before we turn to the extant texts of Cicero, Dionysius and their colleagues, I will now briefly list the four major recurring themes that are at the core of their stylistic discourse. These four leitmotivs will be discussed in depth in the four main chapters of this dissertation (chapters 2–5). In the following sections (1.4 and 1.5), we will see that the surviving discussions of prose style from Late-Republican and Augustan Rome each exhibit some or all of the four central themes. For discussions of the relevant secondary literature for each of these topics, the reader is advised to consult the appropriate chapter.

− Chapter two deals with the selection of models for imitation. In their search for appropriate prose models, the critics and rhetoricians in Rome generally turn to the orators, historians and philosophers of Classical Greece, specifically Athens. Lysias, Isocrates, Demosthenes, Thucydides and Plato are among the most frequently discussed literary giants. As we will see, both Cicero and Dionysius present Demosthenes as the most perfect exponent of artistic prose, while they denounce the Hellenistic historian and orator Hegesias as the ultimate antimodel of good taste.

− Chapter three focuses on various doctrines of three styles. At several points in their oeuvre, Cicero and Dionysius distinguish between three types of style, or stylistic registers ( genera dicendi , χαρακτῆρες τῆς λέξεως ). While Cicero divides style into a plain, intermediate and grand type, Dionysius articulates two three-style divisions, one concerning the selection of words, the other concerning the arrangement of words. Some contemporary authors also recognize three stylistic registers (Varro, Rhetorica ad Herennium ), while others rather identify four styles (Philodemus, Demetrius).

− In chapter four, the art of skillfully combining one’s words will be central. The extant sources for stylistic theory in Rome tend to attach great importance to the issue of word arrangement, or composition ( compositio verborum , σύνθεσις τῶν ὀνομάτων ). In their discussions of the topic, Cicero and Dionysius consistently focus on the aural effects of word arrangement, distinguishing between smooth sounds, that please the ear, and rough sounds, that grate the ear: it is specifically on the aesthetics of ear- grating sounds that this chapter will focus.

CHAPTER ONE

− Finally, chapter five revolves around the ideology of Atticism. Various surviving treatises from Rome insist that prose should be ‘Attic’ ( Atticus , Ἀττικός ), while unsuccessful prose is often condemned as ‘Asian’ ( Asianus , Asiaticus, Ἀσιανός ). As we will see, Atticism is not only concerned with the style of Classical Athenian prose, but it is also connected to such typically Athenian values as wisdom, moderation, freedom and democracy. Thus, by promoting Attic prose, Cicero, Dionysius and others can also make statements about contemporary ethics, politics and society.

It should be noted that the foregoing does not present an exhaustive overview of stylistic themes in the surviving rhetorical and critical treatises from Late-Republican and Augustan Rome. The list could be supplemented, for example, by such topics as the selection of words (ἐκλογὴ τῶν ὀνομάτων , electio verborum ) and figures of speech ( σχήματα; exornationes, lumina, figurae ): after all, many ancient sources divide the topic of style in three parts— selection of words, arrangement of words and figures of speech. 39 Yet, concerning this triad, this dissertation will focus primarily on arrangement, as this was the biggest bone of contention among Greek and Roman scholars in Rome: word selection and figures of speech seem to have been less controversial topics, which will therefore not be addressed in separate chapters. 40 The same goes for the subject of stylistic virtues ( ἀρεταὶ τῆς λέξεως , virtutes dicendi ): following Theophrastus, our Latin sources usually recognize four principal virtues (correct use of Latin/Greek, clarity, ornamentation and appropriateness), while Dionysius uses a complex system of ‘essential virtues’ ( ἀρεταὶ ἀναγκαῖαι ) and ‘additional virtues’ ( ἀρεταὶ ἐπίθετοι ). These various lists of virtues will come up on several occasions throughout this dissertation, especially in my discussion of threefold stylistic divisions (section 3.2). 41

39 Dion. Hal. Isoc. 3.1 (= fr. 691 Fortenbaugh) tells us that Theophrastus enumerated three sources for grandeur: choice of words, their melodious arrangement ( ἁρμονία ), and figures of speech. Such threefold divisions can also be found in the treatment of style in Cic. De or. 3.171–212 and Quint. Inst. orat. books 8–9, distinguishing between ornamentation ‘in individual words’ ( in verbis singulis ) and ‘in combinations of words’ ( in verbis coniunctis ). The latter topic is further divided into word arrangement and figures of speech. Cf. also Long. Subl. 8.1, who lists the three topics as three separate sources for the sublime. Dionysius has not only written a monograph on the arrangement of words, but he also wrote a (lost) treatise on figures and he may have written another treatise on the selection of words: see section 1.5 n. 80 below. Finally, it should be noted that figures of speech are not always treated as a separate category: cf. Demetr. Eloc. 59, Dion. Hal. Comp. 1.8–11, Thuc. 22.1. 40 See however e.g. sections 3.2 and 4.3. 41 De Jonge (2014b) 328–329 offers a brief overview of ancient scholarship on stylistic virtues from Aristotle to Hermogenes; cf. Innes (1985) 255–260, who explores the contribution of Theophrastus to the theory of four

GENERAL INTRODUCTION

The ensuing sections will introduce the surviving Greek and Latin texts that discuss the four chief topics of stylistic theory listed above. I will begin with the principal Latin sources, that is, the rhetorical works of Cicero, the fragments and testimonies of the orator Calvus, and the fourth book of Rhetorica ad Herennium (section 1.4). Next, I will discuss the extant works of contemporary Greek authors who worked in Rome, namely the critical works of Dionysius of Halicarnassus, the fragments of Philodemus’ On Poems and On Rhetoric as well as the fragments of Caecilius of Caleacte (section 1.5). In addition to these texts, I will occasionally use ancillary material from sources that do not strictly pertain to the issue of prose style in Late-Republican and Augustan Rome, such as Latin poetry (Lucilius) and two Greek treatises of uncertain date and origin (Demetrius’ On Style , Longinus’ On the Sublime ).

1.4 Latin Stylistic Theory: Cicero, Calvus and Rhetorica ad Herennium The rhetorical treatises of Cicero are unquestionably the most important sources for the debates about Latin prose style in the first century BC. In fact, Cicero eclipses all other Roman rhetoricians as an authority on stylistic theory, just as he eclipses all other Roman orators as a stylistic model. In this section, we will see that Cicero’s rhetorical works pay close attention to all four major themes of the Greek and Roman stylistic discourse in Rome. Subsequently, we will see that these themes link Cicero’s discussions to the other surviving sources on Latin prose style from the same era. Cicero’s views on prose style can mainly be found in the third book of his dialogue De Oratore (55 BC) and in three later rhetorical works (all 46 BC), namely his dialogue Brutus , his treatise Orator and the brief essay De Optimo Genere Oratorum .42 Cicero’s approach to style in De Oratore differs ostensibly from his treatment of the topic in the later works: while style all but monopolizes Cicero’s attention in 46 BC, it is one of a myriad of topics in his De Oratore . In the latter text, addressed to his brother Quintus, Cicero discusses stylistic theory cardinal stylistic virtues. Cic. Orat. 70 connects the name of Theophrastus to the fourfold scheme. For Dionysius’ system, see Dion. Hal. Thuc. 22–23 and Pomp. 3.8–10. Further relevant literature on the topic is listed in section 3.2.1 n. 36 below: see also ibid. table 4 for a list of essential and additional virtues in Dionysius. 42 For the sections on stylistic theory in De or. , we have Leeman et al. (1996) and (2008), viz., the last two installments of the monumental five-volume commentary on De or. (the first four books are in German, the last one, on De or. 3.96–230, is in English). I have also consulted the annotated translation of May and Wisse (2001) and the commentary of Mankin (2011). For Brut., I have used the German commentary of Kroll and Kytzler (1962) and the English commentary of Douglas (1966). For Orat. , I have used the German commentary of Kroll (1913). See also the notes to the Loeb texts of Brut. and Orat . by Hendrickson and Hubbell (1962), and Opt. gen. by Hubbell (1949).

CHAPTER ONE from the vantage point of the orator’s encyclopedic education: he not only pays due attention to all five tasks of the orator, but he also dwells extensively on the orator’s philosophical training. In the third book, the dialogue’s protagonist L. Licinius Crassus gives a long exposition about style, insisting that stylistic rules are null and void if they are not linked to an all-encompassing knowledge of human life and virtue. 43 The passage about style itself systematically addresses the four cardinal virtues of style—correct Latin (§3.37–48), clarity (§3.49–51), ornamentation (§3.148–209) and appropriateness (§3.210–213). 44 Under the heading of ornamentation, De Oratore successively focuses on the selection of words (§3.149–170), the arrangement of words (§3.171–199) and figures of speech (§3.200–209). The second of these sections, which I have identified as one of the central topics of stylistic theory in Rome, discusses the combining of individual words, prose rhythm and periodic sentence structure. The other three central stylistic topics, on the other hand, are much less prominent in De Oratore . In the dialogue, Cicero pays little attention, for instance, to the selection of appropriate prose models: the effusive praise of Demosthenes, which is an integral feature of Cicero’s later rhetorical works, is conspicuously absent in De Oratore .45 In addition, Cicero barely touches on the doctrine of three styles nor does he distinguish between the Attic and Asian styles of oratory in De Oratore , although these issues are, again, core characteristics of his works of 46 BC. 46 In a word, Cicero pays close attention to word arrangement in De Oratore , but it is in the decade following its publication that he becomes intimately engaged with the three other chief topics of stylistic theory listed above. Indeed, Cicero’s later rhetorical works record a radically different approach to stylistic theory. The dialogue Brutus , to begin with, features Cicero himself as the main interlocutor, exploring the history of Roman oratory with his friends Atticus and Brutus. 47 The

43 See esp. Cic. De or. 3.96–125. The protagonist is the rhetor Crassus (140–91 BC), not to be confused with his grandnephew Marcus, the triumvir Crassus. Marcus Antonius (143–87 BC), the grandfather of the triumvir Marc Antony, features as Crassus’ foil. For a introduction to all interlocutors, see May and Wisse (2001) 14–15. 44 At De or. 3.52–147, Cicero inserts an intermezzo in which Crassus argues that eloquence ( eloquentia ) and wisdom ( sapientia ), which have mistakenly been separated since Socrates, should in fact be united. 45 Cic. De or. 1.58 and 3.71 refers to Demosthenes as a ‘perfect’ ( perfectus ) orator, but on the former occasion he has to share the title with Hyperides, and on the latter occasion with Pericles. 46 Cic. De or. 3.177, 3.199, 3.212 briefly lists three ‘types’ ( figurae ) of style. Cf. section 3.2 below. 47 The dialogue was published in 46 BC before the battle of Thapsus (April 6): see Hendrickson in Hendrickson and Hubbell (1962) 4–5. Recently, there has been a modest surge in scholarly attention for Brut. : see e.g. Fox (2007) 177–208 on Cicero’s skeptical and ironic approach to writing the history of Roman oratory, Stroup (2010) 237–268 on Cicero’s effort to shape the afterlife of Brut ., and Bishop (2016) on Cicero’s veiled criticism

GENERAL INTRODUCTION conversation starts as a lament over the death of the orator Hortensius and grows into an account of the latter’s oratorical ancestors, whose stylistic attributes sometimes prompt more general discussions about style. When Cicero discusses the eloquence of the elder Cato and Calvus, for instance, he adds expositions about the true nature of Attic style and the best models of Attic prose (§63–69 and 283–297). To his treatment of Hortensius’ oratory, alternatively, he adds an excursus on the properties of Asian style (§325–327). Brutus is the first text in recorded history that unconditionally declares Demosthenes the best orator of all time. 48 Cicero repeats his praise in the essay De Optimo Genere Oratorum , which professes to be the introduction to Cicero’s translations of the two opposing speeches of Demosthenes and Aeschines in the trial of Ctesiphon; Demosthenes’ speech is now commonly known as On the Crown .49 In the succinct text, Cicero touches on such issues as the theory of three styles (§1– 6) and the characteristics of Attic style (§7–18), occasionally adding arguments about the relative suitability of Lysias and Demosthenes as stylistic models for oratory in Rome. 50 The treatise Orator , lastly, presents Cicero’s fullest extant treatment of stylistic theory. 51 The text, in which the author sets out to define the ideal orator at the request of Brutus, is almost exclusively devoted to the subtleties of prose style and can be divided into

of Caesar. Dugan (2005) 172–250 discusses a striking irony in the dialogue: Cicero presents the history of oratory as reaching both its acme and its end in his own speeches. Guérin (2014) explores the apparent dissimilarities between Cicero’s division of style in two types at Brut. 201 and the three-style divisions in De or. and Orat. : cf. section 3.4 below. Narducci (2002a) provides a useful overview of the dialogue’s contents and context, as well as a review of the scholarly debate up to 2002. 48 Cic. Brut. 35, 141, 289. On the extreme praise of Demosthenes in the first century BC, see section 2.1 n. 1. 49 Doubts about the authenticity and date of Opt. gen. (e.g. Dihle (1955) 303–314 and Bringmann (1971) 256– 260) have largely subsided: Berry (1996) has demonstrated how Cicero’s use of prose rhythm in the text, which matches that in Brut. and Orat. , is a strong argument in favor of its authenticity. According to Ronconi (1998) 43–68, Cicero left the text in a rough state and he did not finish his translation, because he found a more challenging project in Orat. In addition, Marinone (2004) convincingly shows that Opt. gen. and Orat. must have been composed at roughly the same time. Indeed, as we will see on several occasions, the two works are thematically very closely related: see e.g. section 2.3.2. 50 La Bua (2014) discusses Cicero’s promotion of ‘Demosthenic style’ in Opt. gen. , paying particular attention to the example of his speech Pro Milone that Cicero himself adduces at Opt. gen. 10. 51 Orat. was published briefly after Brut. , as is mentioned at Orat. 23. Recent discussions about the treatise focus on Cicero’s self-fashioning: see esp. Dugan (2005) 251–314 and Bishop (2016). Dugan understands Orat. as the announcement of Cicero’s definitive retreat from public life (esp. Orat. 140–148): cf. however the present section as well as section 5.5.1 below. Narducci (2002b) gives a general overview of the dialogue’s structure and the main topics of discussion in modern scholarship up to 2002.

CHAPTER ONE roughly two parts: the first half is primarily concerned with the doctrine of three styles (§20– 139), while the second half is entirely reserved for the theory of word arrangement (§140– 236). As in De Oratore , Cicero’s discussion of word arrangement includes sections on combining individual words (§149–164), on balanced sentence structure (§164–167) and on prose rhythm (§168–236) respectively. Yet, unlike De Oratore , Cicero now adopts a classicizing perspective to stylistic theory: throughout his analyses, he sets out to identify the key properties of Attic and Asian oratory, and he assesses the merits of Lysias, Demosthenes, Thucydides and other Attic authors as stylistic models. Again, Demosthenes is repeatedly declared to be the best model for artistic prose. 52 In brief, Orator is not only Cicero’s most extensive discussion of prose style that has come down to us, but it also combines all four major recurring themes of the Greek and Roman stylistic discourse. As such, Orator lends itself particularly well for comparison with the stylistic discussions in other sources, especially Dionysius, who, as we will see, also examines all four cardinal stylistic topics. Before we turn to Cicero’s colleagues, however, we should first dwell a little bit longer on the apparent dissimilarities between De Oratore on the one hand and Brutus, Orator and De Optimo Genere Oratorum on the other hand: what has led Cicero to change his approach to style so drastically in the years between 55 and 46 BC? At least two major factors are at play—political circumstances and rhetorical debate. 53 First, Cicero’s rhetorical works incorporate statements about contemporary Roman politics. In 55 BC, Cicero had just returned from exile to find Rome effectively ruled by the triumvirate of Caesar, Pompey and Crassus: by setting his De Oratore in 91 BC, just before the outbreak of the Social War and the civil war between Marius and Sulla, Cicero traces the ideal orator back to a period when speech, in his view, still wielded a formidable power in the political arena. 54 The dialogue ends on an optimistic note: Crassus (the triumvir’s great-uncle) presents the young Hortensius as a brilliant new orator who may go on to fulfill all the requirements for the ideal orator that the interlocutors of De Oratore had stipulated in the course of their conversations. 55 Ironically or not, Cicero still declares himself hopefulabout the future of oratory.

52 Cic. Orat. 6, 23, 26–27, 104–105, 110–111, 133. 53 For the political aspects of Cicero’s shift, see esp. Dugan (2005) 254–267 and Bishop (2016). For the influence of oratorical trends, specifically the rise of Calvus’ Atticist movement, see e.g. Hendrickson (1926) 242–245, Dugan (2001) 409–413, Narducci (2002) 408–412 and Guérin (2014) 168–171. 54 For a general discussion of the political circumstances of the composition of De or. , see Fantham (2004) 1–25. Cf. Dugan (2013) 31: ‘ De oratore offered a nostalgic idealization of the orator.’ 55 Cic. De or. 3.228–230. Cf. section 5.6.1 below on Cicero’s discussion of Hortensius’ style.

GENERAL INTRODUCTION

When he returned to writing rhetorical theory nine years later, it was clear, however, that Hortensius had been unable to deliver on Crassus’ prophecy: every trace of Cicero’s optimism about the fate of eloquence had faded away. 56 In 46 BC, Caesar was the unmistakable master of Rome and Cicero was preoccupied with the tyrannical silencing of free speech: Cicero no longer resorts to happier days for a nostalgic idealization of the orator, bluntly stating that ‘eloquence has become mute’ ( eloquentia obmutuit ). 57 In his Orator , according to Dugan, Cicero acquiesces in the new political reality by focusing mainly on minor stylistic issues such as the intricacies of prose rhythm: ‘Ambitious claims for oratory would ring hollow after the triumph of the force of Caesar’s armies.’58 Yet, as we will see in this dissertation, Cicero’s rhetorical works of 46 BC are not declarations of submission but they rather reflect his preparedness to fight for his ideals (section 5.5.1). This pugnacity is especially reflected in his insistence on the superiority of Attic oratory in general and of Demosthenes in particular: Cicero underscores the importance of Athenian values such as freedom and democracy, while he casts himself as a Roman Demosthenes, rebelling against the tyranny and censorship of Caesar, who is Cicero’s Philip of Macedon. 59 The second aspect that can explain the shift in Cicero’s approach to stylistic theory is the rise to fame of a group of Roman orators, who called themselves ‘Attic’ ( Attici ). Cicero does not refer to these self-styled Atticists in De Oratore , but they feature prominently in his rhetorical works of 46 BC. It is not clear when exactly these orators were active, but the movement seems to have been centered around the orator (and neoteric poet) C. Licinius Macer Calvus, the only name that can be connected beyond doubt to the Atticist movement (section 5.2). 60 The record of Calvus’ activities is limited to the period between 56 and 54 BC;

56 Cf. Dugan (2013) 35: ‘The idealism of De oratore’ s investigation of the ideal orator meets with the hard facts of history.’ Dugan shows that Brut. takes the form of a laudatio funebris , not only for Hortensius, but also for the art of eloquence in Rome. On the evolution of Cicero’s opinion of Hortensius (‘from rivals into partners’), see Dyck (2008). Cf. section 5.6.1 on the role that Hortensius played in the stylistic debates in Rome in the 50s BC. 57 Cic. Brut. 22. 58 Dugan (2013) 38. Cf. Dugan (2005) 253–267. 59 Cf. Bishop (2016) 190: ‘The comparison Cicero drew between Rome under Caesar and Demosthenes’ Athens allowed him to get a pointed political message across: Caesar had put an end to Roman oratory, just as Philip and Alexander had stifled and eventually silenced Athenian free speech.’ In section 5.5.1 below, we will see that it is Cicero’s mission to succeed where Demosthenes had ultimately failed, that is, to salvage democratic eloquence. 60 See section 5.2 below on the origins and the nature of Atticism, which has attracted a vast amount of secondary literature in the late-nineteenth and early-twentieth centuries. As we will see, modern scholarship on Atticism is fraught with misconceptions: with Whitmarsh (1996), I will argue that Atticism was a ‘malleable’

CHAPTER ONE he probably died not much later, certainly no later 47 BC. 61 Two years later, in the summer of 45 BC, Cicero proudly reports that, after the death of their leader, the self-styled Atticists ‘have fallen silent’ ( conticuerunt ), as they were ‘almost jeered out of the Forum’ ( paene ab ipso foro irrisi ). 62 To summarize, in 55 BC Cicero apparently does not yet consider the views of the Atticists worthy of a lengthy refutation, and in 45 BC Cicero no longer considers them an important factor in the public life of Rome: it is fair to assume, therefore, that the peak of their activities lies sometime in the intervening decennium. In his Brutus and Orator , Cicero defends himself against the attacks that Calvus and the so-called Atticists launched against his oratory. Calvus insisted that only speeches that are frugal and devoid of lavish ornamentation are deserving of the title ‘Attic’; he considered Cicero’s extravagant, passionate rhetoric, conversely, quintessentially un-Attic and even ‘Asian’ (section 5.6.1). Naturally, Cicero does not agree with that assessment: he goes to great lengths to show that Calvus ‘was in error and caused others to err with him’. 63 Thus, Calvus’ opinions about prose style are crucial for our understanding of Cicero’s approach to stylistic theory in his rhetorical works of 46 BC. At the same time, these works are the most important surviving sources of the stylistic views of Calvus. From Cicero’s hostile account and several less scathing later sources, we know that Calvus dealt with all four main themes of the stylistic discourse in Rome: 64 favoring the simple rhetoric of Lysias, he was not only involved in the debates about the nature and models of Attic oratory (sections 5.2 and 5.6), but his views also pertain to the theory of stylistic registers (section 3.4) and word arrangement (sections 4.2 and 4.5). Hence, we should understand Calvus not merely as a foil to Cicero, but also as an important participant in the stylistic discourse in Rome.

construct that could be adapted to suit specific purposes. For the historical problems surrounding Calvus and his group of Roman Atticists, see esp. Bowersock (1979) 59–65 and Wisse (1995) 67–69. 61 See esp. Münzer (1926) 428–436, cf. section 5.6 n. 118 below. Calvus is spoken of as dead in Cic. Brut. 283– 284 (early 46) and Fam. 15.21.4 (december 46). According to Douglas (1966) xiii, Calvus died in 47 BC. The consensus, however, is that he died in 54 or 53 BC: see Münzer (1926) 433, Shackleton Bailey (1977) 428–429, Bowersock (1979) 61 and Wisse (1995) 68–69. The most important reason for this date is, in the words of Münzer: ‘Ein Mann mit seinen Fähigkeiten und Leidenschaften, Erfolgen und Aussichten wäre in den nächsten, an Ereignissen reichen und bis in zahllose Einzelheiten wohlbekannten Jahren nicht von der Bühne des öffentlichen Lebens gänzlich verschwunden, wenn er das J. 54 noch längere Zeit überlebt hätte.’ 62 Cic. Tusc. 2.3. 63 Cic. Brut. 284: Et ipse errabat et alios etiam errare cogebat. 64 The main sources for Calvus’ views of style are Cic. Brut. 283–284, Sen. Controv. 7.4.6, Quint. Inst. orat. 9.4.1, 12.1.22, 12.10.12–14 and Tac. Dial. 18.4–5, 21.1–2.

GENERAL INTRODUCTION

A third major Latin voice in the discourse about prose style is the author of Rhetorica ad Herennium , which draws on a similar stylistic framework as Cicero and Calvus. Although date and authorship of the four-volume treatise are still somewhat controversial, Rhetorica ad Herennium is generally assumed to have been written by an unknown author in the mid- to late 80s of the first century BC. 65 The discussion of style, which covers the entire fourth book, begins with a long preface (§4.1–10), in which the author censures Greek authors of rhetorical handbooks for not using their own examples to illustrate stylistic principles, but instead drawing examples from famous orators and poets. 66 Next, the author divides the topic of style in two parts: he first explores the three types of style and the corresponding faulty types (§4.11–16), adding self-written passages to illustrate each of them, and subsequently he expounds the crucial qualities of style, including correct Latin and clarity (§4.17), the proper arrangement of words (§4.18) and figures of speech (§4.19–69). 67 Thus, Rhetorica ad Herennium exhibits two recurring leitmotivs of the Greek and Roman stylistic discourse—the three styles and the theory of word arrangement. As we will see, there are several other Latin sources from Late-Republican and Augustan Rome that can shed some light on the four main themes of contemporary stylistic theory. M. Terentius Varro (116–27 BC), for instance, appears to have used a threefold stylistic division and he was aware of the theory of euphonious word arrangement.68 Additionally, the satirist Lucilius (latter half second century BC) 69 also refers to the art of

65 The work cannot be older than 86 BC, as the author mentions Marius’ seventh consulship and his death in January of that year ( Rhet. Her. 4.68). There is no obvious terminus ante quem , but a date in the mid- or late 80s seems the most plausible on account of (1) the many references in the work to events that took place in that period and (2) the similarities between the work and Cicero’s On Invention , which can be more securely dated to the mid- or early 80s: see e.g. Caplan (1954) xxvi, Adamietz (1960), Achard (1985) and Gaines (2007) 174–177. Still, several scholars have made arguments that Rhet. Her. was published a decade or more later: see e.g. Douglas (1973) and Winkel (1979). Some attribute the anonymous work to a certain Cornificius, who is only known to us through a number of remarks in Quintilian: see e.g. Calboli (1969) 3–11. Yet, the evidence is slight, as is shown by Achard (1989) ix-x. Calboli (1969) offers a commentary to the text. 66 On the author’s (inconsistent) approach to using examples, see Calboli (1969) 46–50. 67 The author treats these topics under three headings: first ‘taste’ ( elegantia ), which includes correct Latin and clarity, secondly ‘arrangement’ ( conpositio ) and thirdly ‘distinction’ ( dignitas ), which discusses figures of speech. These categories correspond to three traditional virtues of style (correct Latin, clarity and ornamentation) —the fourth (appropriateness) is lacking: see Calboli (1969) 301–303. 68 For Varro’s division of three styles, see section 3.3 below. For his views on euphony, see section 4.4 n. 76. 69 For a good discussion of the date of Lucilius’ birth and death, see Herbert-Brown (1999), who follows the dates given at Jer. Chron. 143e and 148e, that is, 148–103 BC.

CHAPTER ONE word arrangement (section 4.2), as does the Augustan poet Horace (65–8 BC), although the latter’s views will not be discussed at length in this study.70 Concerning the notions of Attic and Asian style, to conclude, I will also turn to Suetonius’ report on the oratorical style of the emperor Augustus (63 BC–14 AD), whose sensitivity to the intricacies of prose style suggests a close familiarity with the works of contemporary critics and rhetoricians (section 5.5.2).

1.5 Greek Stylistic Theory in Rome: Dionysius, Caecilius and Philodemus The extant Greek texts on prose style can be linked to contemporary Latin works, listed in the previous section, through the aforementioned main themes of stylistic theory (section 1.3). The present section introduces the three Greek protagonists in the record of the stylistic discourse in Rome: I will mainly focus on the large corpus of critical treatises by Dionysius of Halicarnassus, but I will also touch on the surviving bits and scraps from the works of Caecilius of Caleacte, and the ever-expanding number of fragments of Philodemus of Gadara. Finally, I will briefly discuss Demetrius’ On Style and Longinus’ On the Sublime , that do not (necessarily) originate in Rome, but that nonetheless provide illuminating parallels to the (other) Greek and Latin texts from the city. The vast legacy of Dionysius’ critical treatises, often referred to as his ‘essays’, not only contains invaluable information about ancient stylistic theory, but also about the interrelations between Greek and Roman scholars in Rome. There is no way of knowing when exactly Dionysius published his various critical treatises, but their chronological order can be partly reconstructed: in several passages, Dionysius refers to other works that he has already published. 71 On the basis of Dionysius’ explicit remarks, it is generally accepted that he published his oeuvre in three phases, which represent distinct stages in the evolution of his critical methods: several modern scholars have established that Dionysius’ later works exhibit more sophisticated methods and a larger apparatus of technical vocabulary than his earlier works. 72 It is possible that the Greek teacher gradually became more acquainted with various

70 Chapter 4 focuses on the musical-aesthetic approach to word arrangement: Hor. Ars. p. 46–48 and 240–243, however, refers not to tone and rhythm, but to syntax and semantics. See section 4.1 n. 4. For good discussions of the similarities between Hor. and Dion. Hal. Comp. , see Hunter (2009) 126–167 and De Jonge (2019). 71 On the chronological order of Dionysius’ works, see esp. Bonner (1939) 25–38 and De Jonge (2008) 20–23, who offers an extensive bibliography up to 2008. 72 See esp. Bonner (1939) and Lebel (1973). Schenkeveld (1975a) considers Dionysius’ evaluative methods incoherent, but his view is challenged by Damon (1991), who concludes that the critic’s methods are coherent but incomplete, arguing that we should take the relative order of his works into account. Porter (2016) 221 posits

GENERAL INTRODUCTION grammatical, philosophical and literary doctrines during his stay in Rome: his contacts with the many Greek and Roman scholars in Augustan Rome may have played an important role in the development of his knowledge. 73 By exploring his extant critical works and his known relationships with various Greek and Roman colleagues in Rome, we can partly reconstruct Dionysius’ network and intended audience. Twelve of Dionysius’ critical essays have come down to us. 74 The early phase consists of the treatises On Imitation (which partially survives in fragments and an epitome) 75 and the first part of On the Ancient Orators (which was intended to comprise six separate essays on individual Attic orators), 76 including the famous preface and the essays On Lysias, On Isocrates and On Isaeus .77 The works of the middle period are On Demosthenes (which belongs to the second part of On the Ancient Orators ), On the Arrangement of Words (also known as On Composition ), the Letter to Pompeius (on Plato and the historians) and the First Letter to Ammaeus (on the claim that Demosthenes was influenced by Aristotle’s On Rhetoric ). 78 The later essays, finally, are On Thucydides , the Second Letter to Ammaeus (which is an appendix to On Thucydides ) and On Dinarchus. 79 The treatises On Figures and that modern scholars are too worried about issues of consistency, arguing that Dionysius never meant his categories to be taken ‘in a hard-and-fast way’ and that ‘the various surface inconsistencies in Dionysius’ theory and evaluative practice are best explained by the provisional nature of his schemas’. 73 This has been suggested by Schenkeveld (1983) 69 and De Jonge (2008) 33–34. 74 All surviving critical essays have been edited, with notes, in five Budé volumes: Aujac (1978), (1988), (1991), (1992), Aujac and Lebel (1981). I have also used the notes in the two Loeb editions: Usher (1974) and (1985). 75 Battisti (1997) offers a critical edition, with Italian translation and commentary, of Imit. See now also Schippers (2019), who explores the discourse on imitation in Early-Imperial Rome, comparing Dionysius’ views on the topic in On Imitation to other Greek and Latin discussions of the topic, e.g. Quint. Inst. orat. 10. 76 In Ant. orat. 4.5, Dionysius announces that he will write essays on three orators ‘from the older generation’ ( ἐκ τῶν πρεσβυτέρων ), namely Lysias, Isocrates, Isaeus, all of which survive. He also refers to a second set of essays on three orators ‘from the those who flourished after these’ ( ἐκ τῶν ἐπακμασάντων τούτοις ), that is, Demosthenes, Hyperides and Aeschines. Only a treatise on Demosthenes survives. 77 Hidber (1996) provides an introduction and commentary to the preface of Ant. orat. 78 Ronnet (1952) provides a commentary in French on Dem. Van Wyk Cronjé (1984) focuses on the purpose of Dem. , its place within Ant. orat. and its chronological relation with Comp. For Comp. , we have three editions with introductions, translations and notes: in English Roberts (1910), in French Aujac and Lebel (1981), and in Italian Donadi and Marchiori (2013). See also Pohl (1968), who focuses on the theory of three types of word arrangement ( Comp. 21–24). Roberts (1901) provides an edition, with introduction, translation and notes, of Dionysius’ three extant literary letters. Fornaro (1997) offers an Italian introduction and commentary to Pomp. 79 See the commentaries of Pavano (1958) and Pritchett (1975) on Thuc. , Marenghi (1971) on Din. and Roberts (1901) on Amm. II.

CHAPTER ONE

On Political Philosophy do not survive; it is doubtful if Dionysius ever wrote the essays On Hyperides , On Aeschines and On the Selection of Words. 80 In his surviving works, Dionysius occasionally devotes some attention to the first two tasks of the orator, i.e., invention and disposition, but on the whole he overwhelmingly focuses on style. The four major themes in the stylistic discourse in Rome all receive ample discussion in Dionysius’ critical works. The list of titles already makes it abundantly clear, for instance, that the selection of models is a prominent issue in Dionysius’ approach to style: he not only presents an annotated reading list of Classical Greek literature in On Imitation , but he also addresses the stylistic features of Lysias, Isocrates, Isaeus, Demosthenes, Thucydides and Dinarchus in eponymous treatises. Like Cicero, Dionysius argues that Demosthenes is the most perfect exponent of artistic prose. 81 Next, one cannot fail to notice Dionysius’ obsession with Attic prose: with the exception of Herodotus, Dionysius only discusses orators, historians and philosophers from Classical Athens, who composed their works in the Attic dialect. 82 The preface to On the Ancient Orators is particularly relevant for our understanding of Dionysius’ Atticism. 83 In the brief essay, Dionysius expresses his gratitude to the rulers of Rome for restoring the ‘Attic muse’ ( Ἀττικὴ μοῦσα ) to her former glory. 84 As we will see, Dionysius’ version of the history of rhetoric and style from Classical Athens up to contemporary Rome is remarkably similar to the one that Cicero and other Greek and Roman authors offer (section 1.6). In addition to the early essays of On the Ancient Orators , this dissertation will focus especially on two works from Dionysius’ middle period, viz., On Demosthenes and On the Arrangement of Words . These works were composed at roughly the same time: in the latter

80 Quint. Inst. orat. 9.3.89 refers to Dionysius’ On Figures . Dion. Hal. Thuc. 2.3 mentions his work On Political Philosophy (Ὑπὲρ τῆς πολιτικῆς φιλοσοφίας ), which he had already published. Dionysius announces the composition of On Hyperides, On Aeschines (at Ant. orat. 4.5), and On the Selection of Words (at Comp. 1.10), but there is no proof that he actually proceeded to write these treatises. The Ars Rhetorica that has survived under Dionysius’ name is spurious: see Russell (1979) and Heath (2003). 81 Dion. Hal. Dem. 8.2–4, 14.1, 33.1, 33.4. 82 Note that Dionysius converts the passages from Herodotus that he quotes into the Attic dialect: see Comp. 4.8, Dem. 41.2. See Usher (1974) 398–399: ‘Herodotus was something of an embarrassment to Dionysius’ and ‘he could never be a satisfactory model because he wrote in the Ionic dialect’. Cf. section 1.6 n. 124 below. 83 The preface has been called ‘the manifesto of classicism’ (‘das klassizistische Manifest’) by Hidber (1996). Goudriaan (1989) regards the text as ‘the manifesto of Atticism’ (‘het atticistische manifest’). For the relationship between the terms classicism and Atticism, see section 1.6 n. 120 below. 84 Ant. orat. 3.1. For a more detailed discussion of Dionysius’ praise of Rome, see section 5.5.2 below.

GENERAL INTRODUCTION part of On Demosthenes, Dionysius mentions On the Arrangement of Words , while he refers to a passage from On Demosthenes in the first part of On the Arrangement of Words .85 Therefore, most scholars agree that Dionysius interrupted his work on On Demosthenes in order to write On the Arrangement of Words : in this view, the latter work was published between two separate tranches of On Demosthenes .86 In any case, the two treatises not only originate in the same period, but they are also thematically connected. There are at least two important parallels: both treatises feature threefold stylistic divisions and lengthy discussions of word arrangement. Thus, these texts are extremely valuable for the purposes of this dissertation, as they can be readily compared to Cicero’s Orator , which, as we have seen in the previous section, also focuses on the three styles and the theory of word arrangement. As we will see (section 3.2), Dionysius actually expounds two divisions of three styles —the ‘types of diction’ ( χαρακτῆρες τῆς λέξεως ) and the ‘types of arrangement’ ( χαρακτῆρες τῆς συνθέσεως ). The former division, which is mainly concerned with the selection of words, plays a major role in the first half of On Demosthenes (§1–34). The other threefold division, which concerns word arrangement, can be found in the second half of On Demosthenes (§35– 52) and in On the Arrangement of Words (§21–24). The latter treatise is the only surviving work from Antiquity that is exclusively devoted to the subject of word arrangement: it systematically discusses the nature and force of word arrangement (§2–5), the activities required of the author (§6–9), the aims of word arrangement and the means of attaining them (§10–20), the three types of arrangement (§21–24) and the connections between prose and poetry (§25–26). The treatise combines a wide knowledge of various disciplines, including grammar, rhetoric, philosophy, music and meter. In addition, Dionysius applies the various doctrines of word arrangement to prose as well as poetry, not only quoting passages from oratory and history, but also from lyric (Pindar, Sappho, Simonides), drama (Sophocles, Eurpides, Aristophanes) and epic (Homer). Hence, On the Arrangement of Words is rightly considered Dionysius’ most original contribution to the canon of ancient rhetorical theory. 87

85 In Dem. 49.2 and again in Dem. 50.10, Dionysius refers his readers, who want te learn more about the topic of arrangement, to Comp. In Comp. 18.14, conversely, Dionysius cuts his criticism of Plato’s diction short, because he has given a clear exposition of the topic ‘elsewhere’ ( ἑτέρωθι ): this must be taken to refer to Dem. 5–7, where Dionysius exposes the errors of Plato’s diction. Cf. Van Wyk Cronjé (1986) 97–99. 86 Opinions differ on the precise sequence of publications: see De Jonge (2008) 22–23 for a comprehensive overview of the debate. For the purposes of this dissertation it suffices to know that both Dem. and Comp. were written in roughly the same period. 87 See esp. De Jonge (2008) 41–42.

CHAPTER ONE

We have seen that Dionysius pays due attention to the four hottest issues in the discourse on style in Late-Republican and Augustan Rome. Interestingly, he dedicates his essays to both Greek and Roman intellectuals: these contacts are members of Dionysius’ ‘network’ in Rome. 88 Some of the addressees of his works are only known to us through Dionysius: the Greek Demetrius, for example, to whom Dionysius dedicated his treatise On Imitation , cannot be identified. 89 The same goes for Ammaeus, the addressee of two extant letters as well as the multivolume project On the Ancient Orators , and Pompeius Geminus, who also received a literary letter: we cannot even be sure if Ammaeus and Pompeius were Greek or Roman. 90 We are on firmer ground with the two other known addressees of Dionysius’ work: we know, for instance, that the young Roman aristocrat Metilius Rufus, who received On the Arrangement of Words as a birthday gift, went on to become proconsul of Achaea. 91 The work On Thucydides , lastly, was dedicated to Q. Aelius Tubero, a historian and lawyer, whose two sons both rose to the rank of consul. 92 As noted above (section 1.2),

88 We know the names of several of Dionysius’ contacts in Rome. Roberts (1900) and Hidber (1996) have described these contacts as members of a ‘circle’, but Wisse (1998) rightly objects that this term might wrongly suggest a ‘tightly knit group’, for which there is no evidence in the case of Dionysius and his contacts. Wisse (1995) 78–80 rather uses the more apt word ‘network’, allowing for ‘many contacts, of various sorts and varying intensity, between numerous Greek and Roman intellectuals’; cf. De Jonge (2008) 26 n. 134 and De Jonge and Hunter (2019) 8–9. Wiater (2011) 22–29 uses social identity theory to analyze Dionysius’ network, describing it as an ‘elite community of classicists’. 89 Dion. Hal. Pomp. 3.1, 3.20. The argument of Goold (1961) 178–189 that this man is the same Demetrius as the one who composed Eloc. is unconvincing: cf. n. 106 below. 90 For the dedications to Ammaeus, see Dion. Hal. Ant. orat. 1.1, Amm. 1 1.1, Amm. II 1.1. At Pomp. 1.1, Dionysius tells us that Pompeius received copies of Dionysius’ treatises through their ‘mutual friend Zeno’ (Ζήνων ὁ κοινὸς φίλος ): this suggests that Pompeius was not one of Dionysius’ closer associates and that he was not necessarily in Rome with Dionysius. De Jonge (2008) 27 collects the various hypotheses that have been put forward about the identity of Pompeius. Cf. Hidber (1996) 7 n. 50 on the possibility that Pompeius was a Roman. We have no information about the aforementioned Zeno. 91 Dion. Hal. Comp. 1.4. Dionysius calls Rufus ‘son of a good father who is my most esteemed friend’ ( πατρὸς ἀγαθοῦ κἀμοὶ τιμιωτάτου φίλων ). In addition to the fact that Dionysius includes the Metilii in his list of Alban leaders ( Ant. Rom. 3.29.7), this may imply that Rufus’ father acted as Dionysius’ patron. Bowersock (1965) 132 n. 2 identifies Rufus with the Metilius who became proconsul of Achaea and perhaps legate of Galatia. Dionysius offers Comp. to Rufus ‘as he is celebrating the first birthday of his manhood’ ( πρώτην ἡμέραν ἄγοντι ταυτηνὶ γενέθλιον ). Weaire (2012) studies Dionysius’ strategies as a Greek teacher of a Roman student in Comp. 92 Dion. Hal. Thuc. 1.1, cf. Amm. II 1.1. In Ant. Rom. 1.80.1, Dionysius mentions Tubero’s historical work. See Bowersock (1965) 130 and Bowersock (1979) 68–69. De Jonge (2008) 28 is reluctant to accept Bowersock’s thesis that Tubero was Dionysius’ ‘patron’.

GENERAL INTRODUCTION

Cicero knew both Tuberones (son Quintus as well as father Lucius), which makes it all the more plausible that Dionysius was (directly or indirectly) familiar with Cicero’s works. It seems, then, that Dionysius intended his critical works to be read by all civilized men, Greeks and Romans alike. 93 Hence, it is appropriate that he is often described as an author standing between Greece and Rome. 94 A similar picture emerges from his Roman Antiquities , in which he emphasizes the connections between Greek culture and the early Romans. 95 In the first book of his history, Dionysius claims that he learnt Latin and that he read the works of various Roman historians, including Varro. 96 For the present study of Greek and Roman stylistic theory, I take away two crucial points. First, Dionysius employs an apparatus of theories, techniques and terminology that was widely used by Greek as well as Roman scholars in Late-Republican and Augustan Rome. Secondly, Dionysius was actively involved in the public life of the city and had ample opportunity to exchange ideas and doctrines with his colleagues. On these two accounts, Dionysius’ critical essays lend plausibility to my thesis that Greek and Roman scholars in Rome not only participated in a shared stylistic discourse, but that they were also engaged in a mutual dialogue about the style of artistic prose. To the list of Dionysius’ known acquaintances we should add the name of the Sicilian rhetorician Caecilius of Caleacte (fl. Augustan era), whom Dionysius calls a ‘very dear friend’. 97 According to the Suda , Caecilius was a ‘Jew clever in Greek matters’, who taught rhetoric in Rome and composed numerous treatises: he was not only acquainted to Dionysius,

93 On the intended audience of Dionysius’ critical works, see De Jonge (2008) 23–25. In most cases he seems to write for scholars who are already well-versed in the study of rhetoric: cf. Dem. 46.4, Thuc. 25.2. The only exception is Comp. , which is primarily addressed to his young pupil Metilius Rufus, but which also caters more generally ‘for young men and those who are just beginning to take up their study’ ( τοῖς μειρακίοις τε καὶ νεωστὶ τοῦ μαθήματος ἁπτομένοις ὑμῖν). 94 E.g. Luraghi (2003): ‘Dionysios von Halikarnassos zwischen Griechen und Römern’. See also De Jonge and Hunter (2019) 2: ‘A writer positioned between Greece and Rome’. 95 We should note that there is a vehement debate about the intended audience of his Ant. Rom. : see De Jonge and Hunter (2019) 31–33. Wiater (2011) 118 argues that Dionysius intended to separate Greek and Roman audiences, perhaps even provoking his Roman readers: cf. the review in De Jonge (2012b), who rightly emphasizes ‘the bridge that Dionysius builds between Greeks and Romans’. 96 See Ant. Rom. 1.6.2, 1.7.3. In Ant. Rom. 1.14.1, he mentions Varro’s work Antiquities (Ἀρχαιολογίαι ). 97 Dion. Hal. Pomp. 3.20 (= T36 Woerther = XIV 163 Ofenloch): τῷ φιλτάτῳ Καικιλίῳ. Tolkiehn (1908) points out that the word φίλτατος is rarely used in Dionysius’ work and that it hence signifies a close connection with Caecilius. On the faulty dating of Caecilius by Wilamowitz (1900) 15 and Nassal (1910) 9, see section 1.2 n. 12 above.

CHAPTER ONE but also to the rhetoricians Hermagoras of Temnus and Timagenes of Alexandria, who both taught rhetoric in Rome. 98 Although only a few fragments and some titles survive, it is obvious that Caecilius took a special interest in Attic oratory and its models, i.e., two of the four central themes of the stylistic discourse in Rome. 99 He wrote the (lost) works On the Stylistic Character of the Ten Orators , On Lysias and On Demosthenes as well as a Comparison between Demosthenes and Aeschines and a Comparison between Demosthenes and Cicero : Caecilius is known to have criticized Plato and to have praised Lysias and Demosthenes, while his discussion of Cicero shows that his interest was not limited to Greek literature alone. 100 Among his other known works are treatises entitled How the Attic and Asian Styles Differ and Against the Phrygians , which probably was an alphabetically arranged lexicon of ‘Asian’ diction. 101 Even from this incomplete overview of Caecilius’ lost works, it

98 Suda κ 1165: Ἰουδαῖος σοφὸς τὰ Ἑλληνικά (= T1 Woerther). The text gives us Archagathus as his original name and notes that he may have been of servile parentage. Roberts (1897) gives a good summary of the available evidence for Caecilius’ activities. For the association with Hermagoras and Timagenes, see Suda ε 3024 and τ 588 (T2–3 Woerther), and Woerther (2015) 49–50. 99 The three subsequent editors of Caecilius have become increasingly hesitant in attributing fragments and testimonies to him: the standard edition of Ofenloch (1907) adopts the most inclusive approach, allowing various texts that do not mention Caecilius by name as sources for his works. Augello (2006), conversely, only includes texts that explicitly refer to Caecilius, while Woerther (2013) is even more rigorous in avoiding spurious or dubious attributions to the rhetorician. Recent discussions of Caecilius’ stylistic views focus on his role in the creation of canons of writers (see section 2.3 n. 34 below) and on his conception of sublime literature, for which see esp. Innes (2002) and Porter (2016) 184–195. 100 Suda κ 1165 (= T1 Woerther): Περὶ τοῦ χαρακτῆρος τῶν δέκα ῥητόρων , Περὶ Δημοσθένους , Σύγκρισις Δημοσθένους καὶ Αἰσχίνου , and Σύγκρισις Δημοσθένους καὶ Κικέρωνος . For ‘his work in defence of Lysias’ ( τὰ ὑπὲρ Λυσίου ), see Long. Subl. 32.8: ‘While he loves Lysias more than he loves himself, he hates Plato more than he loves Lysias’ ( φιλῶν γὰρ τὸν Λυσίαν ὡς οὐδ’ αὐτὸς αὑτόν , ὅμως μᾶλλον μισεῖ Πλάτωνα ἢ Λυσίαν φιλεῑ). Yet, Caecilius also criticized Lysias: cf. T44 Woerther = VII 110 Ofenloch. Anastassiou (1966) 33–34 saw that it is impossible to ascertain Caecilius’ preferred model; Innes (2002) 276–288 thinks that he favored Demosthenes; Porter (2016) 189–193 speculates that he attached himself in particular to Aeschines. The method of comparing two or more authors ( σύγκρισις ) was also a favorite method of Dionysius ( Pomp. 1.6–8): see Focke (1923), Vardi (1996) and De Jonge (2018b) for the importance of this critical method, cf. also section 2.3.1 n. 42 below. 101 Suda κ 1165 (= T1 Woerther): Τίνι διαφέρει ὁ Ἀττικὸς ζῆλος τοῦ Ἀσιανοῦ and Κατὰ Φρυγῶν. About the latter, Suda informs us that it is ‘alphabetically arranged’ ( κατὰ στοιχεῖον ). The next book on Suda ’s list is Demonstration that Every Word of Elegant Language has Been Spoken (Ἀπόδειξις τοῦ εἰρῆσθαι πᾶσαν λέξιν καλλιρρημοσύνης ), also alphabetically arranged: this may have been a lexicon of Attic vocabulary, a counterpart of Against the Phrygians. For Phrygia as a symbol of Asianism in Dion. Hal. Ant. orat. 1.7 and Cic. Orat. 25, see section 1.2 n. 36 above.

GENERAL INTRODUCTION is clear that the rhetorician was actively involved in the debate about the Attic and Asian styles of oratory, and that he took an interest in both Classical Greek and recent Latin literature: in these respects, he stands on equal footing with his friend Dionysius, but also with his Roman predecessors Cicero and Calvus. 102 For the third major Greek source of stylistic theory in Rome, we must go back in time from the Augustan to the Republican era, to the versatile Epicurean philosopher and epigrammatist Philodemus of Gadara (ca. 110–after 40 BC), who was active in Rome and Herculaneum from the mid-70s onwards. 103 His works are now gradually being recovered from the carbonized papyrus scrolls that were found in the ‘Villa dei Papiri’ in Herculaneum. Philodemus’ aesthetic works are particularly relevant to us because of their extensive discussions of word arrangement: the topic features in some surviving bits and pieces of On Rhetoric and, more prominently, in the fragments of the five-book treatise On Poems .104 The latter work can be read as a passionate diatribe against Crates of Mallos and various Hellenistic critics (now standardly referred to as οἱ κριτικοί ), who stressed the importance of euphony and word arrangement in the evaluation of poetry. 105 Additionally, On Poems is also relevant to the evaluation of prose: after all, as Cicero and Dionysius assert, both poetry and prose apply tone and rhythm in the arrangement of words (section 4.5). Finally, we should note that Philodemus associated himself with various prominent Romans during his stay in

102 We should also mention Caecilius’ On Figures (cf. T13–27 Woerther = IV Ofenloch), which yielded the most citations in later sources, and On the Sublime (cf. T28–31 Woerther = V Ofenloch), to which Longinus reacts in his treatise by the same name (see the present section below), and which is linked to the doctrine of stylistic registers, as we will see in section 3.2 below. 103 The available information about Philodemus’ life is summarized by Janko (2000) 4–7. The date of his death is somewhat ambiguous: the terminus post quem is 40 BC, as Philodemus ( On Signs 3 col. 2,18) refers to the pygmees that Marc Antony brough to Rome in that year. Philodemus probably died sometime in the 30s BC. 104 The international ‘Philodemus project’, led by David Blank, Richard Janko and Dirk Obbink, gradually publishes Philodemus’ aesthetic texts, with translations and notes. Thus far, Poem. 1 and Poem. 3–4 have come out: Janko (2000) and Janko (2011). Volumes containing Poem. 5 (Armstrong, Porter, Fish and Mangoni), Rhet. 1–2 (Blank), Rhet. 3 (Blank) and Rhet. 8 (Obbink and Hammerstaedt) are being prepared. For Poem. 5 we may consult the edition of Mangoni (1993) with the translation of Armstrong (1995b). The standard edition for Rhet. is Sudhaus (1892–1896), on which Hubbell (1920) bases his partial translation. Large sections of Sudhaus’ edition (and the book order that he adopted) have now been replaced: see Longo Auricchio (1977) for Rhet. 1–2 and Hammerstaedt (1992) for Rhet. 3. 105 Modern scholars on ancient literary criticism usually study the text of Poem. for the views of o ἱ κριτικοί and the so-called ‘euphonist tradition’ to which they are reckoned to belong: see esp. Schenkeveld (1968), Porter (1995), Janko (2000) and section 4.1 below.

CHAPTER ONE

Italy: Cicero praises his learning, he was a protégé of Caesar’s father-in-law L. Calpurnius Piso Caesonius and he knew several famous Latin poets, including Virgil and perhaps also Horace. 106 To conclude this section about the Greek sources for stylistic theory, we should mention two works that cannot be unequivocally connected to Late-Republican and Augustan Rome. The first is the treatise On Style by a certain Demetrius (perhaps second or early first century BC), 107 which is thematically linked to the discourse on prose style in Rome: the author recognizes four types of style, which he distinguishes according to the criteria of substance, selection of words and arrangement of words. The work includes several passages on euphony, rhythm and periodic sentence structure (esp. §1–35, 38–74, 179–184, 204–208 and 241–271). 108 The last work to be discussed is On the Sublime by an unknown author, who is conventionally called (pseudo-)Longinus (perhaps first century AD). 109 This work takes the form of a polemic against the ‘little treatise’ on the same topic by Caecilius of Caleacte. 110

106 For Cicero’s references to Philodemus, see Pis 68–72 and Fin. 2.119. In the former text, Cicero tells us that Philodemus was Piso’s constant companion (cf. Asc. Pis. 68). Philodemus also knew the consul C. Vibius Pansa Caetronius, to whom he dedicated Rhet. 4. He was a friend of Siro, in whose Neapolitan house Virgil studied Epicurean philosophy; Philodemus dedicated his work On Vices and Virtues to Virgil and the young poets P. Quinctilius Varus, L. Varius Rufus and Plotius Tucca. Horace cites an epigram of Philodemus at Serm. 1.2.121, but it is uncertain if the two men ever met: the case for direct contact is convincingly presented by Oberhelman and Armstrong (1995) 235–236. For the influence of Philodemus on Horace, see also Brink (1963) 43–152. 107 For good recent discussions on the date of Eloc. (περὶ ἑρμηνείας ), see Innes in Halliwell et al. (1995) 310– 321 and, more concisely, De Jonge (2009). While Grube (1961) has an earlier date (ca. 270 BC) and Marini (2007), following earlier scholarship, proposes a later date (first century AD), there is growing consensus that the work belongs to the second or early first century BC: see esp. Chiron (1993) and Innes in Halliwell et al. (1995). Schenkeveld (1964) 135–148 thinks that Eloc. is a reworking from the first century AD of material reflecting the second or early first century BC; Schenkeveld (2000), however, places the work firmly in the first century BC. 108 I have consulted the rich commentary of Marini (2007) as well as the notes of Roberts (1902) and Chiron (1993). Schenkeveld (1964) contains valuable studies of Demetrius’ stylistic theories. 109 The name Longinus comes from the only manuscript for Subl. , which has Διονυσίου Λογγίνου in its title and Διονυσίου ἢ Λογγίνου in the table of contents. For a brief overview of the controversy about authorship and date, see Russell in Halliwell et al. (1995) 145–148. Whereas Heath (1999) thinks that the work was written by Cassius Longinus in the third century AD, most scholars nowadays put it in the mid- to late first century AD: see esp. Russell (1964) xxii-xxx, Häussler (1995) and Innes (2002) 259. According to Goold (1961), Mazzucchi (2010) and De Jonge (2014a), the treatise might even belong to the Augustan era. 110 Long. Subl. 1.1: ‘The little treatise by Caecilius, which he wrote on the sublime’ ( τὸ τοῦ Καικιλίου συγγραμμάτιον , ὃ περὶ ὕψους συνετάξατο ). I have used the commentaries of Russel (1964) and Mazzucchi (2010) on Subl.

GENERAL INTRODUCTION

The author addresses various issues that will sound familiar to the readers of Cicero and Dionysius: On the Sublime not only pays much attention to such Attic authors as Plato and Demosthenes, comparing the latter to Cicero (§12), but it also includes sections on the procedure of imitating these sublime models (§13–14) and on the topic of word arrangement (§39–42), which the author declares to be one of the five principal sources of sublimity in poetry and prose: he claims to have written a separate two-book treatise on the topic. 111 In this dissertation, both On Style and On the Sublime will provide illuminating parallels to the material from Cicero, Dionysius and their Greek and Roman colleagues in Rome.

1.6 Athens into Rome: the Shared Classicism of Cicero and Dionysius Having established not only the capital themes of stylistic theory in Rome (section 1.3), but also the relevant Greek and Latin sources on these recurring topics (sections 1.4 and 1.5), I will no cursorily explore the classicism that underlies the shared discourse on prose style in Late-Republican and Augustan Rome: the topic of classicism merits a separate discussion in this introductory chapter, as it pertains to our understanding of the dialogue between Greeks and Romans in the city. The present section will show that Cicero, Dionysius and their colleagues, despite their obvious differences, adopt broadly the same approach to the history of Greek oratory and the study thereof in contemporary Rome. In a word, they exhibit similar conceptions of classicism. There exists no comprehensive, unambiguous definition of the term ‘classical’: it can nowadays be used to describe anything from Homer’s poems to Augustan Rome and Polyclitus’ Doryphorus, not to mention Mozart’s symphonies or Newton’s laws of motion. 112 The ancients themselves barely use the word ‘classical’ at all: at no point, for instance, do Cicero and Dionysius explicitly refer to any author, style or era as such.113 Still, it is justified

111 Long. Subl. 39.1: ‘On this matter (i.e., word arrangement) I have in two books given a sufficient account of such conclusions as I could reach’ ( ὑπὲρ ἧς ἐν δυσὶν ἀποχρώντως ἀποδεδωκότες συντάγμασιν , ὅσα γε τῆς θεωρίας ἦν ἡμῖν ἐφικτά ). There do not survive any fragments of the works nor any testimonies about its contents. 112 See Porter (2006b) for an excellent introduction to the incoherence of ancient and modern definitions of the classical: ‘It is important to realize that there were no unified views about questions of classical value in Antiquity, any more than there have been in modernity.’ 113 The modern use of the word ‘classical’ goes back to Fronto, cited in Gell. NA 19.8.5: ‘A first-class, authoritative writer, not one of the common herd’ (classicus assiduusque aliquis scriptor, non proletarius ). The word classicus invoked not only a military connotation (a classis is a fleet of warships), but also social-economic status, as becomes clear in Gell. NA 6.13.1, quoting the elder Cato: ‘Not all those men enrolled in the five classes were called classici , but only men of the first class, who were rated at a hundred and twenty-five thousand asses

CHAPTER ONE to subsume their views under the heading of ‘classicism’, as they both apply a threefold division of history, which is an integral feature of most, if not all views, tendencies and products that can be identified as ‘classicistic’ or ‘classicizing’: tripartite periodizations not only appear in the stylistic discussions of men like Cicero and Dionysius, but also in later movements, such as the Second Sophistic, Renaissance painting and Neoclassical architecture. Basically, all these ‘classicists’ divide the history of art and literature into the following three phases (what Thomas Gelzer calls the ‘klassizistische Dreischritt’): first an idealized golden age in the distant past, then a period of decline, and lastly a revival of the golden age in the present. 114 Needless to say, this crude scheme is extremely flexible and open for continuous renegotiation: in Antiquity, the ‘classical past’ was most often situated in Athens in the fifth and fourth centuries BC, but classicists through the ages disagreed about the essence of Athenian genius and about the appropriate approach to renewing its glory.115 Thus, classicism was not a recognized school of thought in Antiquity (it did not even have a name), but the term may be adduced to shed light on ancient debates about art and literature. Art historians, for example, have for a long time studied the effects of classicism in the visual arts in Greece and Rome. 116 In recent decades, moreover, there has been a considerable surge in scholarly attention for the interplay of classicism and identity among Greek authors living under Roman rule: while it is traditionally understood as a defense of Hellenic civilization against the domination of Rome, Tim Whitmarsh has convincingly or more’ ( classici dicebantur non omnes, qui in quinque classibus erant, sed primae tantum classis homines, qui centum et viginti quinque milia aeris ampliusque censi erant ). Cf. Citroni (2006) 204–211, who argues that Gellius’ scriptor classicus does not refer to an exclusive list of emblematic authors, but rather to a large group of authors considered to be reliable authorities. 114 For the phrase ‘klassizistische Dreischritt’, see Gelzer in the discussion of Preisshofen (1979) 278; for the theories and observations underlying Gelzer’s coinage, see, in the same book, Gelzer (1979) 3–13. The phrase is also used by Hidber (1996) 14–25 in his discussion of Dion. Hal. Ant. orat. , which he, as we have seen (n. 82 above), describes as a ‘manifesto of classicism’ (‘klassizistisches Manifest’). 115 Cf. Porter (2006b): ‘We are evidently having to do not with a single form of classicism but with a variety of classicisms in the plural, each differently conceived, some less ‘purely’ than others, and more often than not in polemical dialogue with contemporaries and predecessors, among whom the very standards of purity will be up for debate.’ 116 The modern interest in ancient expressions of classicism goes back to the work of the German art historian and archeologist Winckelmann (1717–1768). More recent influential explorations of classicism in Greek and Roman art include Pollitt (1972), who focuses on the ancient conceptions of the Greek visual arts of the High- Classical period, and Zanker (1988) 239–264, who discusses how in Augustan Rome classicism and archaism in art and architecture were part of a conscious program to sustain the emperor’s transformation of Roman society.

GENERAL INTRODUCTION shown that Greek imperial classicism is too complex and variegated to be reduced to a single model. 117 The conceptions of the classical in Latin literature have been the subject of fewer studies, although Mario Citroni has published a series of articles on the conceptions of the classical exemplified in the development of Latin literary canons. 118 In this dissertation, I am primarily interested in the interconnections between Greek and Latin theories in Late- Republican and Augustan Rome.119 We should now, therefore, zoom in on Cicero and Dionysius, starting with their threefold divisions of the history of rhetoric. In the works of Cicero and Dionysius, the three successive stages in the history of rhetoric are not only distinguished chronologically but also geographically: the classical era is conflated with Athens, the following dark age with Asia Minor, and the ultimate rebirth with the city of Rome. Hence, the terms ‘Atticism’ and ‘Asianism’ can be used to further qualify the classicism in the stylistic discourse of Cicero, Dionysius and their fellow critics and rhetoricians (sections 5.1 and 5.2). 120 In his introduction to On the Ancient Orators , Dionysius offers a clear articulation of his tripartite periodization; I quote the opening lines here.121

117 Most studies of Greek classicism focus on the Second Sophistic (ca. 50–ca. 250 AD). Swain (1996), for instance, suggests that the expressions of Greek classicism in this period can often be interpreted as assertions of Greekness directed against Rome. Whitmarsh (2001), (2013a), (2013b) et alib., however, emphasizes the ambivalent approach of the Greeks to their Roman conquerors, interpreting the various expressions of Greek classicism as personal, ingenious responses to the socio-political reality. Several scholars have also paid attention to classicism in the preceding era: see esp. the collections of essays in Flashar (1979) on Rome in the first centuries BC and AD, and Porter (2006a) on classicism throughout the centuries. For Dionysius’ classicism, see esp. Goudriaan (1989), Hidber (1996), De Jonge (2008) 9–20 and Wiater (2011), cf. section 1.2 n. 25–27, 77. 118 See e.g. Citroni (1998), (2003) and (2006): following Gelzer (1979), he submits that the ancient concept of the classical has two separate meanings, i.e., the axiological sense of exemplary excellence and the typological sense of conformity to certain formal aesthetics, such as measure, organic structure, balance, idealization and nobility: in the author’s view, the latter interpretation in Latin is articulated for the first time in the Augustan age. 119 Bowersock (1979) and Wisse (1995) discuss some links between Roman and Greek views of Atticism, focusing on the possible routes of transmission, which will not concern us here (cf. section 1.2 item 1b above). 120 Classicism and Atticism are not identical, as Gelzer (1979) 13–14 explains: Atticism is a specific manifestation of the wider phenomenon of classicism. Incidentally, the terms Atticism and Asianism are also, like classicism, modern coinages; see section 5.2 below on the flexibility of the categories ‘Attic’ and ‘Asian’. 121 Dion. Hal. Ant. orat. 1.1–2. This passage is partly quoted and discussed in section 5.5.2 below.

CHAPTER ONE

κρείττω . Ἐν γὰρ δὴ τοῖς πρὸ ἡμῶν χρόνοις ἡ μὲν ἀρχαία καὶ φιλόσοφος ῥητορικὴ προπηλακιζομένη καὶ δεινὰς ὕβρεις ὑπομένουσα κατελύετο , ἀρξαμένη μὲν ἀπὸ τῆς Ἀλεξάνδρου τοῦ Μακεδόνος τελευτῆς ἐκπνεῖν καὶ μαραίνεσθαι κατ ’ ὀλίγον , ἐπὶ δὲ τῆς καθ ’ ἡμᾶς ἡλικίας μικροῦ δεήσασα εἰς τέλος ἠφανίσθαι .

We ought to acknowledge a great debt of gratitude to the age in which we live, my most accomplished Ammaeus, for an improvement in certain fields of serious study, and especially for the considerable revival in the practice of civil oratory. In the epoch preceding our own, the old philosophic rhetoric was so grossly abused and maltreated that it fell into a decline. From the death of Alexander of Macedon it began to lose its spirit and gradually wither away, and in our generation it had reached a state of almost total extinction.

Dionysius connects the downturn of rhetorical excellence to the death of Alexander the Great (323 BC): this date is retained to this day to mark the watershed between the Classical and Hellenistic eras. The Greek critic uses an elaborate allegory to describe the opposite styles of oratory on either side of this critical moment (section 5.4): he refers to the ‘old, philosophic rhetoric’ as the ‘Attic muse’ ( Ἀττικὴ μοῦσα ), while its vile successor was ‘an upstart that had arrived only yesterday or the day before from some Asian death-hole, a Mysian or Phrygian or a Carian creature’ ( ἔκ τινων βαράθρων τῆς Ἀσίας ἐχθὲς καὶ πρῴην ἀφικομένη , Μυσὴ ἢ Φρυγία τις ἢ Καρικόν τι κακόν ).122 In Dionysius’ own day, as he himself claims, a change for the better occurs, for which he thanks his host city (section 5.6.2): ‘I think that the cause and origin of this great revolution has been the conquest of the world by Rome ( ἡ πάντων κρατοῦσα Ῥώμη ), which has made every city focus its entire attention upon her.’ 123 In a word, Dionysius celebrates the literary genius of Classical Athens, which had faded away in the Hellenistic era and would have been completely wiped out ‘in my own generation’ ( ἐπὶ δὲ τῆς καθ ’ ἡμᾶς ἡλικίας ), if it were not for the civilizing effect of Rome’s dominion. 124

122 Dion. Hal. Ant. orat. 1.5–7. The passage is quoted in full and discussed at more length in section 5.4 below. On Mysia, Phrygia and Caria as a shorthand formula for Asianism, see section 1.2 n. 36–37 above. 123 Dion. Hal. Ant. orat. 3.1: Αἰτία δ’ οἶμαι καὶ ἀρχὴ τῆς τοσαύτης μεταβολῆς ἐγένετο ἡ πάντων κρατοῦσα Ῥώμη πρὸς ἑαυτὴν ἀναγκάζουσα τὰς ὅλας πόλεις ἀποβλέπειν . For a more extensive discussion of how Rome had contributed to the revolution according to Dionysius, see section 5.5.2 below. 124 The prose authors whom Dionysius deems worthy of discussing at length are all from the fifth and fourth centuries BC; with the exception of Herodotus, they all write in the Attic dialect (cf. section 1.5 n. 82). The authors whom Dionysius expressly marks as inferior come from the early third century BC (Hegesias of

GENERAL INTRODUCTION

Dionysius does not specify the starting point of the classical era. He seems interested in all Greek literature written before the death of Alexander: among the poets, for instance, he admires Homer, Pindar and Sappho, who predate the Attic orators whom he admires.125 Cicero does not appear to disagree: according to him, the origins of classical style are obscure through a lack of written records.126 Like Dionysius, the Roman author adopts a tripartite periodization of the history of rhetoric, to which he refers in his Brutus and Orator.127

Nam ut semel e Piraeo eloquentia evecta est, omnis peragravit insulas atque ita peregrinata tota Asia est, ut se externis oblineret moribus omnemque illam salubritatem Atticae dictionis et quasi sanitatem perderet ac loqui paene dedisceret. Hinc Asiatici oratores non contemnendi quidem nec celeritate nec copia, sed parum pressi et nimis redundantes; Rhodii saniores et Atticorum similes.

For when once eloquence had sailed forth from Piraeus it traversed all the islands and visited every part of Asia, but in this process it contracted some stain from foreign ways and lost that wholesomeness, and what one might call the sound health, of Attic diction; indeed it almost unlearned the art of natural speech. From this source came the Asian orators, not to be despised whether for their readiness or their abundance, but redundant and lacking conciseness. The school of Rhodes however retained more sanity and more similarity to the Attic source.

While Dionysius’ Attic muse was maltreated and driven away from her rightful home, Cicero imagines that eloquentia simply sailed away from the port of Piraeus. In accordance with his Greek colleague, he sets the date of this journey roughly around the death of Alexander.

Magnesia, Duris, Hieronymus), the late third century BC (Chrysippus, Phylarchus, Psaon, Demetrius of Callatis, Antigonus) and the second century BC (Polybius, Heraclides, Hegesianax). On Hegesias, see esp. sections 2.2 and 2.4 below; on the other authors, see Dion. Hal. Comp. 4.14–22 with Aujac and Lebel (1981) 203–204. 125 See e.g. Dion. Hal. Comp. 20.8–22, 22.11–33, 23.10–17. Cf. De Jonge (2008) 340 n. 62: ‘For writing prose, Attic was the model, but for poetry the dialect depended on genre requirements. Therefore, Dionysius quotes not only Homer, but also Sappho and Pindar in their own dialect.’ Cf. also Comp. 4.14, where Dionysius notes that ‘almost all the ancient writers’ ( οἱ ἀρχαῖοι ὀλίγου δεῖν πάντες ) before the Hellenistic era made a careful study of word arrangement, which according to him separates good from bad authors (cf. ibid. 4.13). 126 Cic. Brut. 27 claims that the written oratorical records began in the time of Pericles and Thucydides, ‘who do not belong to the cradle but to the maturity of Athens’ ( qui non nascentibus Athenis sed iam adultis fuerunt). 127 Cic. Brut. 51 is quoted here. Cf. esp. Orat. 24–27.

CHAPTER ONE

According to Cicero, it was actually the deaths of the last Attic orators—Lycurgus (324 BC), Demosthenes (322 BC), Hyperides (322 BC), Aeschines (314 BC) and Dinarchus (ca. 291 BC)—that marked the institution of ‘certain other types of speaking, which were softer and more relaxed’ ( alia quaedam dicendi molliora ac remissiora genera ), cultivated above all by the orators from Asia Minor. 128 Cicero bluntly claims that, on her travels through Asia, oratory ‘almost unlearned the art of natural speech’: a parallel view can be found in the contemporary Roman grammarian and tragic poet Santra (fl. mid-first century BC), who submits that ‘as Greek gradually extended its range into the neighboring cities of Asia, people who had not yet secured sufficient command of the language acquired a passion for eloquence, and so began to express by periphrases what could have been said directly, and then continued in the same vein’.129 Yet, we should note that Cicero is not as scathing about the Asian orators of the Hellenistic era as Dionysius: he not only points out their vices but also the virtues of their oratory, such as ‘readiness’ (celeritas ) and ‘abundance’ ( copia ). 130 Remarkably, Cicero next presents Rhodes as a golden mean between Athens and Asia, perhaps building on the tradition which held that the Attic orator Aeschines, after losing the trial of Ctesiphon against Demosthenes, founded a rhetorical school on the island; we do not know, however, if there existed a distinct Rhodian style of oratory.131 In my view, both

128 Cic. De or. 2.94–95: ‘As long as imitation of these men persisted, the pursuit of this type of speaking lived on. But when, after their death, all memory of them had gradually grown dim and then vanished, certain other types of speaking flourished, which were softer and more relaxed’ ( Quorum quam diu mansit imitatio, tam diu genus illud dicendi studiumque vixit; postea quam exstinctis his omnis eorum memoria sensim obscurata est et evanuit, alia quaedam dicendi molliora ac remissiora genera viguerunt ). Transl. May and Wisse (2001). Cicero connects this style explicitly to the Asian orators Menecles and Hierocles of Alabanda. Cic. Brut. 37, 285; Orat. 92–96 adduces Demetrius of Phalerum (fl. ca. 300 BC) as a transitional figure between Athens and Asia, declaring him thoroughly Attic but already displaying a florid style like the Asian orators: cf. section 3.3 below. 129 Quint. Inst. orat. 12.10.16–17: Paulatim sermone Graeco in proximas Asiae civitates influente nondum satis periti loquendi facundiam concupierint, ideoque ea quae proprie signari poterant circumitu coeperint enuntiare ac deinde in eo perseverarint. This observation, attributed to Santra, may serve specifically to explain the puffiness, redundancy and lack of conciseness that are often associated with Asian oratory (cf. secion 5.6.1 below). For the surviving fragments and testimonies of Santra, see Mazzacane (1982). 130 On Cicero’s ambivalent approach to Asianism, see esp. Cic. Brut. 325–327 and section 5.6.1 n. 143 below. 131 For Rhodes as a mean between Attic and Asian oratory, see esp. Cic. Brut. 316, Orat. 25, Quint. Inst. orat. 12.10.18–19. For Aeschines’ alleged influence on Rhodian oratory, see Quint. Inst. orat. 12.10.19, [Plut.] X orat. 840d–e, Philostr. VS 481. Portalupi (1957) aims to describe the characteristics of ‘Rhodianism’ (120–70 BC): the presence of such famous rhetoricians as Theodorus of Gadara, Hermagoras of Temnus and Apollonius Molon to some extent justifies the hypothesis that there existed a Rhodian rhetorical tradition. Yet, Douglas (1958) rightly

GENERAL INTRODUCTION

Cicero’s moderate praise of Asia and his reference to Rhodes serve primarily as a defense for his own Greek rhetorical education, which follows a path that neatly matches the itinerary of the ship of eloquence that we have seen above: Cicero relates that he first spent six months in Athens, that he next ‘traveled through all of Asia Minor’ ( Asia tota pegrinata est ), after which he completed his training on Rhodes. 132 Of course, Cicero’s tutelage in Asia made him vulnerable to charges of Asianism: by stressing the importance of his subsequent stay on Rhodes, he can stress his affinity with Attic oratory without disowning the lessons of his Asian teachers. On Rhodes, his tutor Apollonius Molon ‘made it his task to repress if possible the redundancy and excess of my style, which was marked by a youthful impetuousness and lack of restraint, and to check it so to speak from overflowing its banks’. 133 Cicero’s own tour as well as the journey of eloquentia eventually lead back to Rome, where the young orator sets out to bring about the revival of Attic oratory. He was not the only orator in the city to lay claims to this feat: Calvus and his followers already called themselves ‘Attic’ on account of their Lysianic simplicity (section 5.5.1), while there were others who imitated the style of Thucydides or Xenophon. 134 Cicero, however, claims that it is in his own oratory that the stylistic variation of Athens is approximated for the first time: ‘We found the ears of the city hungry for this varied type of oratory, displayed equally in all styles, and we were the first (nos primi ), however poor we may have been and however little we may have accomplished, to turn them to an amazing interest in this style of oratory.’135 What is objects that ‘Atticism and Asianism were not movements with a history but rather catchwords of literary controversy incapable of exact definition’: defining Rhodianism as lying in between these vague polemical extremes is very problematic. On the fluidity of Atticism and Asianism, see esp. section 5.2 below. 132 Cic. Brut. 315–316. Note the almost verbatim repetition of ‘every part of Asia was visited’: peregrinata tota Asia est (Brut. 51), Asia tota peregrinata est (Brut. 315). 133 Cic. Brut. 316: Is dedit operam, si modo id consequi potuit, ut nimis redundantis nos et supra fluentis iuvenili quadam dicendi impunitate et licentia reprimeret et quasi extra ripas diffluentis coerceret. Cf. Douglas (1966) 286: ‘Cicero’s description of Rhodes was nothing more than a compliment to his teachers.’ Stroup (2007) 28: ‘As much as the Brutus passage professes to give a simple overview of the birth and rise of eloquentia in the east, it serves mainly—and importantly—to parallel eloquentia ’s juvenile travels to the orator’s own youthful course of study and present (in 46 BCE) political and social straits.’ 134 For the orators following of Thucydides and Xenophon, see Cic. Orat. 30–32. Cicero objects that this procedure leads to a style that is unsuitable for civic oratory: cf. section 4.5.3 below. 135 Cic. Orat. 106: Ieiunas igitur huius multiplicis et aequabiliter in omnia genera fusae orationis auris civitatis accepimus easque nos primi, quicumque eramus et quantulumcumque dicebamus, ad huius generis audiendi incredibilia studia convertimus. On variation as typical for Attic oratory, see esp. Cic Brut. 285–291 with section 2.2 below. See also section 3.3 on the connections between classicism and the theory of three styles.

CHAPTER ONE more, Cicero insists that it is only in his day (and especially in his own writings) that the Romans discovered prose rhythm and periodic structure, making it possible for the first time to come close to imitating the literary styles of Demosthenes and Isocrates.136 Like Dionysius, then, Cicero situates Rome’s revolutionary return to the heydays of Athens in his own lifetime—in his own prose, to be more precise. The foregoing has shown that Cicero and Dionysius, despite their apparent differences, adopt a similar basic conception of the history of rhetoric. We may safely assume that their perspective was widely shared among contemporary Greek and Roman authors in Rome, be it that others do not articulate their views about the three-period history as clearly as Cicero and Dionysius. We have seen, for instance, that among the Romans Santra made a kindred observation about the transition from Attic to Asian oratory, while Calvus and his followers presented themselves as Attic orators incarnate. As for the Greeks, we may adduce the testimony of Strabo, who also remarked on the end of the Attic golden age and the beginning of Asian degeneracy. 137 Caecilius of Caleacte, lastly, was likewise fascinated by the distinction between Atticism and Asianism, drawing up alphabetic lists that may have served as practical tools for literary composition in Rome. Thus, although this dissertation will reveal numerous points of disagreement between the critics and rhetoricians in Rome, we may conclude that there is ample common ground to be found in their conceptions of the classical.

1.7 The Dialogue between Greek and Latin Stylistic Theory The common classicism described in the previous section underscores one of the principal theses of this dissertation, i.e., that Greek and Roman authors alike tap into a shared repertoire of theories, techniques and terminology in their discussions of prose style (section 1.1). What can these remarkable similarities teach us about the relationship between Greek and Latin stylistic theory in Rome? In this section, I will show that the traditional terminology of

136 Cic. Orat. 171: ‘In fact, nearly four hundred years have passed since this became the approved style in Greece, but we have only recently recognized it’ (et apud Graecos quidem iam anni prope quadringenti sunt, cum hoc probatur; nos nuper agnovimus ). According to Cic. Orat. 175, Gorgias was the inventor of balanced sentence structure, and Thrasymachus was the inventor of prose rhythm: as both men were active in the second half of the fifth century BC, Cicero’s reference to four hundred years makes sense, if we count back from 46 BC. See Orat. 226 on Cicero’s own contribution to the Roman discovery of prose rhythm: ‘We have written more about rhythmical prose than anyone before us’ ( plura de numerosa oratione diximus quam quisquam ante nos ). 137 Str. Geogr. 14.1.41: see section 2.2 n. 14 below. Strabo asserts that the orator Hegesias of Magnesia (fl. early third century BC) initiated the Asian style of oratory: thus, Strabo’s date for the watershed between the Classical and Hellenistic era is compatible with Cicero’s and Dionysius’ propositions.

GENERAL INTRODUCTION acculturation is ill-suited to analyze the relationship between Greek and Latin stylistic theory in Rome. Next, I will demonstrate that the idea of a cultural ‘koine’, which is currently in vogue among archaeologists and historians studying the material culture of the Mediterranean world, offers a convenient framework to understand the points of convergence as well as the apparent differences between our Greek and Latin sources. An obvious way to look at the common ground between Greek and Latin stylistic theory is through the conceptual lense of acculturation: in this approach, the shared discourse on prose style may be seen as the result of some form of contact or exchange between Greek and Roman culture. 138 As noted before, scholars have paid much attention to the Greek influences on Roman ideas, or to the Roman appropriation of Greek ideas (section 1.2): while this hellenocentric perspective may be helpful in studying the diachronic connections between earlier Greek theorists and later critics and rhetoricians, it is unsuitable for analyzing the synchronic relationship between Greek and Latin theories in Rome. 139 Assuming that there was a mutual exchange of ideas between Greek and Roman authors, we should turn to a more dialectical conception of their common stylistic discourse: modern scholarship on cultural contact supplies such metaphors as metallurgical ‘fusion’ (‘two things fuse to create a third which, though blended from others, is completely new’) or biological ‘hybridity’ (‘cross- fertilization of different species creates offspring which are genetically different from both parents but retain characteristics of both’) to understand such mutual influence. In these images, the categories of Greek and Roman become obsolete as a result of their interaction. 140

138 For ‘acculturation’ as an umbrella term covering a wide range of views about the interaction between two or more cultures, see Veyne (1979) 4: ‘Décidément l’acculturation n’est qu’un mot, qui désigne les conséquences variées et subtiles de situations historiques qui sont autant d’intrigues variées et compliquées. A vrai dire, l’acculturation est un phénomène incessant et universel.’ 139 Much work has been done on the ambivalent Roman responses to Greek culture: see e.g. Woolf (1994), who argues that ‘Roman responses to Hellenism consisted of a complex and partly incoherent mixture of adoption, adaptation, imitation, rejection and prohibition’, and Stroup (2007), who regards the influence of Greek rhetoric on Rome as a story of expansion, resistance and acculturation (i.e., appropriation) successively. Again, these thoughtful studies are valuable in establishing the general Roman attitudes to Greek learning, but they do not address the interaction between Romans and Greeks working in Rome. For the Greek literature of the Roman empire as a reaction to Rome, see section 1.6 n. 117 above and the literature cited there. 140 See Wallace-Hadrill (2008) 3–28 for a cogent argument against the traditional terminology of acculturation (fusion, hybridity, creolization, ‘métissage’) in favor of the model of bilingualism, or ‘code-switching’ (see below): ‘The principal stumbling block thus appears to be the tacit assumption that culture is unitary, that you must be one thing or another, or even a blend of the two, but not both at the same time.’ See Newsome (2011) 68 for the descriptions of fusion and hybridity quoted above.

CHAPTER ONE

The critics and rhetoricians in Rome, by contrast, participate in a shared discourse about style without compromising their identity as Greeks or Romans. 141 What we need, then, is a model that accounts for the mutual influence of Greek and Roman stylistic views, whilst maintaining the integrity of the Greek and Roman identities of their authors. The work of Andrew Wallace-Hadrill is helpful in this respect: ‘Culture does not respond to the food blender: you cannot throw in chunks of Greek and Roman, press a button, and come out at the end with a homogeneous suspension of bland pap.’ 142 In his groundbreaking book Rome’s Cultural Revolution , he argues therefore that romanization and hellenization are not mutually exclusive, but rather ‘two closely interrelated aspects of the same phenomenon’, that is, cultural interaction. 143 Building on the notion of bilingualism and the sociolinguistic concept of ‘code-switching’, he shows that the inhabitants of the Roman world could consciously alternate not only between the Greek and Latin language (e.g. the use of Greek in Cicero’s letters), but also between Greek and Roman behaviors and customs (e.g. wearing a toga on the Forum, but a pallium in your private villa). 144 Thus, Wallace-Hadrill

141 On several occasions, the authors of our sources explicitly present themselves as Romans or Greeks. Cic. Orat. 152, for example, distinguishes between ‘Greeks’ (Graeci ) and ‘us’ (nos ). Cf. Rhet. Her. 4.1–10 on several differences between Greek and Roman approaches to style, and Long. Subl. 12.5 referring to the Romans as ‘you’ ( ὑμεῖς) and to Demosthenes as ‘our man’ ( ὁ ἡμέτερος ). Throughout his Ant. Rom. , Dionysius reflects on the connections between Greeks and Romans: cf. Peirano (2010), who shows that Dionysius paints a nuanced picture, often portraying the people of early Rome as ‘Hellenized Romans’, but at the end of the work depicting them rather as ‘barbarized Greeks’. 142 Wallace-Hadrill (1989) 164. This review of Zanker (1988) is entitled ‘Rome’s Cultural Revolution’ and plants the seeds for Wallace-Hadrill’s later work on the interaction between Greek and Roman culture, culminating in Wallace-Hadrill (2008), by the same title as the original review: see n. 143 below. 143 Wallace-Hadrill (2008) 26. The work aims to describe a ‘social and cultural revolution’ in the Late- Republican and Early-Imperial Roman world, that according to the author is at least as dramatic as the ‘political revolution’ that Syme (1939) described. Its most important innovation lies in the idea that hellenization and romanization are related notions that can be compared to the two phases of the pumping of blood through the human body: the heart in this image is Rome, which draws Greek culture into Italy (the ‘diastolic phase’), and next pumps away the subsequent culture into the provinces (the ‘systolic phase’). The cardiovascular analogy rightly stresses the circularity of cultural contact, allowing us to move past a strictly chronological model of Greek influence on Rome: cf. Newsome (2011). One of the metaphor’s shortcomings is that it seems to present Rome as ‘some mighty automaton’: cf. Osborne and Vout (2010) 240 and Versluys (2015) 164–165. 144 On bilingualism and linguistic code-switching in the ancient world, see esp. Adams, Janse and Swain (2002). Wallace-Hadrill (1998) applies the concept of code-switching to material culture, e.g. domestic architecture, where rooms decorated in Greek and Roman styles could be deliberately juxtaposed. For the social and cultural implications of wearing a toga or a pallium in the Roman world, see Wallace-Hadrill (2008) 38–70.

GENERAL INTRODUCTION demonstrates how ‘two cultures may sit alongside one another, interacting rather than assimilating’. 145 Due to the intensive contacts between Greeks and Romans in Republican and Imperial Rome, both groups had at their disposal a wide cultural repertoire, from which they could select appropriate elements to suit particular contexts or purposes. While the concept of code-switching in itself has little to offer to my analysis of the Greek and Latin discussions of prose style, it is worthwhile to further push the notion of a common repertoire underlying these texts. The abundant scholarship on material culture and identity provides us with further tools to do so: archaeologists and historians use the word ‘koine’ (κοινή ) to refer to a widespread ‘repertoire of material culture at hand’ in a given period.146 Studying the complexities of cultural interaction in the Hellenistic and Roman Mediterranean, Miguel John Versluys explains how such a koine may be built up: ‘Through increasing connectivity, all kinds of elements (for the Hellenistic repertoire: mostly Greek, but also Egyptian, Etruscan, Persian, etc.) can be found in ever-widening contexts, acquiring new meanings there and changing in character.’ 147 A similar point, mutatis mutandis, can be made about stylistic theory in Rome: Greek and Roman authors refer to a shared corpus of classical models, while they also draw on a common set of analytical approaches theoretical doctrines from various sources, such as Peripatetic scholarship (esp. Aristotle, Theophrastus), Hellenistic criticism ( οἱ κριτικοί ) and Roman oratory (e.g., the views of the elder Cato, the claims of the self-styled Attic orators in the city) which are constantly rearticulated and reinterpreted. Thus, the works of Cicero, Dionysius and their colleagues adhere to a sort of rhetorical-critical koine, ‘a common language used and supplied and further developed by all participants’. 148

145 Wallace-Hadrill (1998) 84. 146 Versluys (2013) 429. The application of the term koine in scholarship on cultural identity evokes an analogy with ‘koine Greek’ ( κοινὴ διάλεκτος ), the common form of the Greek language of the Hellenistic era: Colvin (2011) 39 shows that ‘the koine constitutes a standard to which no spoken or written variety corresponds exactly’. Likewise, in the analysis of types and styles of material culture, the word koine refers to a widely available common ‘language’ or repertoire whose individual local applications differ considerably, although they take their elements from the same reservoir. Scholars have described such ancient cultural ‘koinai’ for various times: cf. Beaujard (2010) for the Bronze-Age Aegean, Vlassopoulos (2007) for the Classical-era Greek poleis, and Versluys (2013) and (2015) for the Hellenistic and Roman eras after ca. 200 BC. 147 Versluys (2015) 155. Cf. ibid. 158–163 for the argument that from ca. 200 BC onwards ‘a koine of shared cultural symbols is present and functioning all around the oikoumene , and that, hence, we cannot but understand the cultural system as a globalised one from that period onwards’. 148 Versluys (2015) 154.

CHAPTER ONE

To further explain the functioning of such a shared koine, Versluys applies the terms ‘universalization’ and ‘particularization’: ‘Univerzalisation indicates that styles and elements that originally belonged to a specific culture are detached from that specific culture in order to play a role in a larger system. (…) With this common repertoire of universal (global) elements available, we subsequently see particularization in local contexts. Elements from the Hellenistic koine were only made to work and got their specific interpretation through contextual application.’149 The so-called First Style (ca. 200–80 BC) and Second Style (ca. 80–20 BC) of Pompeian wall painting can illustrate how this works. These styles were not only found in Pompeii and Rome, but they were popular all over the Mediterranean: the local applications of the styles are all unique, although they use elements and decorations from a shared reservoir. 150 In Greek and Latin discussions of prose style, likewise, the common (i.e., universal) theories, techniques and terminology can be applied in various ways by the respective critics and rhetoricians according to their specific (i.e., particular) preferences, purposes and programs. To sum it up, the vocabulary of koine, universalization and particularization provides an elegant and convenient framework to understand both the similarities and the differences between the views of Cicero, Dionysius and their fellow critics and rhetoricians. In my view, the most important upside of this approach is that it allows us to focus simultaneously on the overarching trends in the theories and doctrines of a given period as well as on the personal goals and motivations of the individual authors: in this way, we can review the works of ancient scholars within the larger context of their time, while doing justice to their various programs and idiosyncracies. Moreover, the history of rhetoric and criticism is not presented as a linear process, with later authors continually developing the

149 Versluys (2015) 155. In his study of the globalizing effects of Roman imperialism, Witcher (2000) 216–218 adopts the notions of universalization and particularization from scholarship on globalization in the modern world, allowing him to explain the variety among so-called ‘romanized’ regions in the Mediterranean: ‘Roman imperialism led to a series of (re-)negotiations, both local and global, which led to a (re-)definition of societies and identities.’ 150 See Bilde (1993) for a comparison of First-Style paintings at Pompeii with the contemporary cultural koine. See Versluys (2015) 155–156 on the Second Style: ‘Each time, forms and materials from all over the Hellenistic world were combined in novel ways. Nowhere, therefore, do Second Style wall decorations look the same and there are certainly differences between (local) particularizations over the (global) Mediterranean. But the elements are all taken from the same reservoir. Through their particularization, moreover, new combinations are made that, in their turn, are added to the koine with the possibility of being particularized (etc.). Through processes of universalization and particularization the Second Pompeian Style is globalized and globalizing.’

GENERAL INTRODUCTION views of their predecessors, but rather as following a cyclical pattern, i.e., a recurrent cycle of universalization and particularization. First, a term, notion or doctrine is used in a particular context (say, Aristotle’s Rhetoric ); next, it can be adopted by an ever expanding group of users, in the course of which it gradually gets detached from its original context and connotations (i.e., it becomes ‘universalized’); it can thence be adapted to suit new contexts (i.e., it becomes ‘particularized’), after which the novel interpretations can subsequently end up in the koine again and the process starts anew. 151 Admittedly, the procedure described here still offers a rough generalization of the ancient dialogue on style: in the individual case- studies of this dissertation, we will have better opportunities to see in more detail how the Greek and Roman stylistic discourse may work.152 The present section has aimed to provide an answer to the following question: how are the Greek and Latin stylistic theories of Late-Republican and Augustan Rome related to each other? I have argued that a plausible answer to this question can be articulated by assuming that Greek as well as Roman authors had access to a common rhetorical-critical koine, from which they selected and adapted appropriate elements to suit their various objectives. On the basis of this notion of a common language of stylistic theory, we can understand how the

151 This cyclical process can be illustrated, for instance, by tracing the history of the maxim in Arist. Rh. 3.8.1 that ‘the form of style should be neither metrical nor unrhythmical’ ( τὸ δὲ σχῆμα τῆς λέξεως δεῖ μήτε ἔμμετρον εἶναι μήτε ἄρρυθμον ). Aristotle’s statement, which may itself be derived from earlier theorists, can be understood in the context of his quest, in all aspects of life and the arts, for an appropriate mean between two faulty extremes: cf. section 3.5 n. 112 below. In the first century BC, Cic. Orat. 195, 220 adjusts Aristotle’s ‘universalized’ thesis, so as to suit his defense of his own highly rhythmical style: speech, in his view, may be ‘rhythmical’ ( numerosus ), but it may not ‘consist of definite rhythms’ ( e numeris constare ). Although Cicero has abandoned Aristotle’s notion of the golden mean, he still links the view about rhythm and meter in prose explicitly to Aristotle. Dion. Hal. Comp. 11.25, next, presents great prose as being neither ‘in rhythm’ (ἔνρυθμος ) nor ‘unrhythmical’ ( ἄρρυθμος ) without referring specifically to Aristotle: in arguing that prose may even exhibit near-verses, as long as it is not entirely ‘in rhythm’, Dionysius’ view is closer to Cicero’s than to Aristotle’s. In the first century AD, the longstanding doctrine appears to have been wholly detached from its Aristotelian context: see, e.g., Theon Prog. p. 71.7-11 Spengel (see section 2.4.2 n. 141) and Quint. Inst. orat. 9.4.56. In the latter text, we can see that the ‘particularized’ interpretations of Cicero en Dionysius have become part of the koine to which the Flavian teacher has access: Quintilian paraphrases Cic. Orat. 220 by applying the same distinction as Dionysius, viz., ‘in rhythm’ ( enrythmus ) versus ‘unrhythmical’ ( arrhythmus ). Cf. section 2.3.3 n. 89 and section 4.3 n. 52 below. 152 The dialectical development of the stylistic discourse on the basis of both Greek and Roman contributions will be demonstrated, e.g., with respect to the evaluation of rough word arrangement (section 4.5.1) and the various notions of Atticism (section 5.2).

CHAPTER ONE various critics and rhetoricians, with their disparate personal agendas, might engage in a dialogue about the style of prose, dialectically negotiating and renegotiating their shared framework. This approach to ancient stylistic theory helpfully frames the three main theses of this dissertation (section 1.1): it accounts for the availability of a common critical discourse in Rome (thesis 1), for the flexibility of its various elements (thesis 2), and for the active participation of both Greek and Roman authors in it (thesis 3). In the ensuing chapters, I will explore the principal elements of this stylistic koine, starting with Cicero’s and Dionysius’ assessment of the Attic model Demosthenes and the Asian antimodel Hegesias of Magnesia. Focusing on the universal and the particular, we will see that the Roman rhetorician and the Greek critic build their judgments of these authors on a common basis, which they exploit to suit their own purposes.

Chapter 2 HERO VERSUS ZERO : DEMOSTHENES AND HEGESIAS AS PARADIGMS OF GOOD AND BAD STYLE

2.1 Introduction In the foregoing introductory chapter we saw that the extant stylistic discussions from Late- Republican and Augustan Rome generally exhibit an enthusiastic admiration for the authors of Classical Athens, while these same texts often present the authors of the Hellenistic period as inherently inferior. Although the surviving stylistic discussions refer to various Attic models and Hellenistic antimodels (section 2.2), the present chapter will focus on the two authors that Cicero and Dionysius of Halicarnassus identify as their favorite and least favorite authors respectively. We will see, for one, that they both single out the Athenian orator Demosthenes as the best prose author who surpasses all other exponents of Greek prose, whereas they select the Hellenistic orator and historian Hegesias of Magnesia as the very worst writer in the history of literature. The purpose of this chapter is to establish what motivates Cicero and Dionysius to nominate Demosthenes as the epitome of literary perfection and Hegesias as the paradigm of bad taste. Thus, we will be able to see what Cicero and Dionysius consider to be the distinctive qualities of good and bad literature. The figures of Demosthenes and Hegesias offer us convenient shortcuts to the stylistic discourse of Late-Republican and Augustan Rome: Cicero and Dionysius not only analyze Demosthenes and Hegesias in superlative terms, setting them up as incontestable icons of good and bad style, but they also apply a similar theoretical apparatus in their praise and censure. What is more, the focus on Demosthenes and Hegesias not only characterizes the stylistic discussions of Cicero and Dionysius, but it is actually an integral feature of the Greek and Roman stylistic discourse in which they participate. The extreme praise for Demosthenes, to begin with, seems to be a novelty in the first century BC (section 2.3). 1 Cicero, the first

1 Anastassiou (1966) connects the ‘uneingeschränkte Anerkunnung des Demosthenes’ to the second half of the first century BC. Cooper (2000) argues that Peripatetic philosophers equated Demosthenes’ political dishonesty with oratorical dishonesty, which influenced the reception of his speeches in the Hellenistic era: according to Cooper, it is in Late-Republican and Early-Imperial Rome that Demosthenes’ reputation as a stylistic icon was established. Cf. Pernot (2006) 61–62, who shows that the ancient appreciation for Demosthenic oratory (or the lack thereof) is inextricably bound up with the appreciation or depreciation of his behavior as a man and as a politician. For Demosthenes’ political reputation, see section 2.2 below. Cf. also sections 5.6.1 and 5.6.2 below on the importance of Demosthenes as a political model for Late-Republican and Augustan Rome.

CHAPTER TWO surviving source to declare him the indisputable champion of Greek prose, describes him as ‘the perfect orator who lacks absolutely nothing’, saying that ‘it is amazing how much one man has pre-eminence over all’. 2 Dionysius, next, claims that ‘he uses judiciously the best style, and the one which is most perfectly adapted to all aspects of human nature’. The critic pits passages from Demosthenes against specimens from other authors such as Lysias, Thucydides, Plato and Isocrates, concluding that the latter writers ‘are not worthy to compete for the palm with Demosthenes’. 3 The name of Hegesias, by contrast, is used as a virtual synonym for bad taste. He was native to the Lydian town of Magnesia on the river Sipylus and he probably flourished in the early third century BC. 4 Only a few fragments of his extensive literary output survive: most of these either come from his history of Alexander the Great or from his epideictic speeches. 5

2 Cic. Brut. 35: Nam plane quidem perfectum et quoi nihil admodum desit Demosthenem facile dixeris. Cic. Orat. 6: Admirabile est quantum inter omnis unus excellat. 3 Dion. Hal. Dem. 33.1: ( Ἡ πρόθεσις ἦν μοι καὶ τὸ ἐπάγγελμα τοῦ λόγου ) κρατίστῃ λέξει καὶ πρὸς ἅπασαν ἀνθρώπου φύσιν ἡρμοσμένῃ μετριώτατα Δημοσθένη κεχρημένον ἐπιδεῖζαι . Ibid. 33.5: ( Ἴνα δείξαιμι ) Δημοσθένει γε οὐκ ἀξίους ὄντας ἁμιλλᾶσθαι περὶ τῶν ἀριστείων . Dionysius considered the method of comparing authors ( σύγκρισις ) the best method of assessing literature ( Pomp. 1.6–8); Caecilius also used it (section 1.5 above): on σύγκρισις as a critical tool in the exegesis of Demosthenes, see section 2.3.1 below. 4 Hegesias himself gives us the name of his hometown at Dion. Hal. Comp. 4.11 (= F17 Jacoby). It is not possible to give more than an approximate date for his floruit : see Radermacher (1910) 2607–2608 and Staab (2004) 127. There are two main clues for a date in the early third century BC: Str. Geogr. 14.1.41 declares him the instigator of the Asian style of rhetoric (cf. n. 23 below), and Cic. Brut. 286 reports that he aimed to imitate Charisius, an Attic orator contemporary to Menander and Demetrius of Phalerum, who flourished around 300 BC and whose fame seems to have been limited to his own lifetime: cf. Blass (1880) 318, and n. 27 below. 5 In referring to the fragments and testimonies of Hegesias, I use the numbering of Jacoby (1929), Fragmente der griechischen Historiker IIB 142. Piotrowicz (1925), Jacoby (1930) and Prandi (2016) supply commentaries on the surviving texts. There is one substantial fragment (F5) on Alexander’s siege of Gaza (332 BC), which is preserved in Dion. Hal. Comp. 18.26 as a sample ‘from his history’ ( ἐξ ἱστορίας ), but no details about the work are known: see Prandi (2016). Spina (1989) compares Hegesias’ version of the siege to other sources; Donadi (1999) discusses the problems in the transmission of the text. The other fragments are much shorter, varying from one to a few sentences: it is often unclear whether such fragments belong to his historiographical works or to his speeches: cf. section 2.4.1 n. 98. He is often called a rhetor (T1, T7, F30–31 Jacoby) or a sophist (F5 Jacoby). It is likely that he composed epideictic speeches: he is credited with a Praise of Rhodes (ἐγκώμιον Ῥόδου , F1–2 Jacoby). There is some evidence that he wrote judicial speeches: Rutilius Lupus, On Figures of Speech 1.7 (= F27 Jacoby) contains an address to a jury, but this could also be a school exercise, as Jacoby (1930) 529 points out. Hegesias possibly also composed dialogues, although only three titles survive: a Rhodian inscription (= T11 Jacoby) attributes The Lovers of Athens (Οἱ φιλαθήναιοι ), Aspasia and Alcibiades to Hegesias.

HERO VERSUS ZERO

Virtually all of his extant words are preserved in hostile contexts as examples of bad taste: his first appearance is in the work of the historian and geographer Agatharchides of Cnidus (second century BC), but, just as Demosthenes receives his highest praise in the first century BC, Hegesias receives most of his scorn in that era as well (section 2.4). Although Varro is said to have approved of his style, Cicero claims that ‘anyone acquainted with him needs to look no further for an example of ineptitude’. 6 Dionysius diagnoses him with mental illness and depicts him as a ‘high-priest of humbug’. In addition, Dionysius predicts that ‘one could not find a single page that is felicitously composed’ in the entire corpus of Hegesias’ writings. He claims that Hegesias’ words ‘are likely to be uttered only by women or emasculated men, and not even by them in earnest, but in the spirit of mockery and ridicule’. 7 The gist of the matter seems to be that Demosthenes does virtually everything right and that Hegesias does almost everything wrong. While the surviving stylistic analyses of Demosthenes’ and Hegesias’ works do not reveal much about the actual hallmarks of their styles, they do offer us valuable insights into the stylistic views of the critics. 8 In this chapter, I will first explore the apparent preoccupation with Demosthenes and Hegesias against the background of the dominant stereotypes of Athens and Asia in Rome (section 2.2). Next, I will focus particularly on the stylistic analyses in the works of Cicero and Dionysius: what is it exactly that they like about Demosthenes’ style (section 2.3) and what is it that they dislike about Hegesias’ style (section 2.4)? By answering these questions, I will be able to make inferences about the shared literary tastes of Cicero and Dionysius, and about the shared

6 Cic. Orat. 226: (Hegesias’ style is so bad) ut non quaerat quem appellat ineptum qui illum cognoverit. In Att. 12.6.1, Cicero produces a parody of the ‘style of Hegesias, which Varro commends’ ( genus Hegesiae, quod Varro laudat ): cf. section 2.4.2 below for a rhythmical analysis of this pastiche. About ‘Asian’ features in Varro’s style, see Traglia (1982). Philodemus and Rutilius Lupus (late first century BC) mention Hegesias without an obvious negative assessment: see section 2.2 n. 25 on Philodemus and section 2.4 n. 94 on Rutilius. 7 Dion. Hal. Comp. 18.21: Διαφθορὰ τῶν φρένων . Ibid. 4.11: Τῶν λήρων ἱερεύς . Ibid. 18.23: Μίαν οὐκ ἂν εὕροι τις σελίδα συγκειμένην εὐτυχῶς. Ibid. 18.28: Ὡς δὲ ὁ Μάγνης εἴρηκεν , ὑπὸ γυναικῶν ἢ κατεαγότων ἀνθρώπων λέγοιτ ’ ἂν καὶ οὐδὲ τούτων μετὰ σπουδῆς, ἀλλ ’ ἐπὶ χλευασμῷ καὶ καταγέλωτι . 8 Indeed, it has been observed that Cicero’s and Dionysius’ observations are often not pertinent to the works that they analyze: see e.g. Wooten (1989), who argues that Dion. Hal. fails to grasp the essence of Demosthenes’ style . Wooten (1997) 177–187 likewise argues that Cicero’s criticisms of Demosthenes are not valuable for students of Demosthenes: cf. section 3.4 below. According to Wiater (2008) 111–114, Dionysius’ criticism of Hegesias is not based on any real study of his texts, but only on prejudice. Cf. section 3.1 below on the limited usefulness of the ancient three-style theories for the stylistic analysis of Greek and Latin prose, and section 4.4 below on the applicability of Dionysius’ theories of composition and euphony. Vaahtera (1997) shows that Dionysius’ theories of euphony often do not work for the passages that he cites: see esp. section 4.4 n. 68 below.

CHAPTER TWO critical framework on which they build their arguments (section 2.5). It should be noted that this chapter is chiefly concerned with the similarities between the views of Cicero and Dionysius: it aims to present a synopsis of their common approach to good and bad prose. The relevant differences between their stylistic views, specifically with respect to Demosthenes’ oratory, will be dealt with in the ensuing chapters of this dissertation.

2.2 Demosthenes and Atticism, Hegesias and Asianism The prominence of Demosthenes and Hegesias in the stylistic discussions of Cicero, Dionysius and their contemporaries must be understood as a concomitant of the preoccupation of these scholars with the classicist categories of Atticism and Asianism (section 1.6). Demosthenes was associated more than any other prose author with the literary, cultural and political legacy of Classical Athens, or, as Cicero puts it, ‘I believe that not even Athens itself was more Attic than he.’ 9 Hegesias, on the other hand, was reviewed as a prototype of the degenerate literature, culture and politics of Asia. Thus, Demosthenes and Hegesias were not only singled out on the basis of their stylistic attributes, but also on the basis of their suitability as cultural and political icons. In this section, I will examine how Demosthenes and Hegesias came to be regarded as the archetypes of Attic and Asian style respectively on the basis of the biases about Athens and Asia that were current in Late-Republican and Augustan Rome. In the first century BC, the city of Rome saw an increased focus on the Athenian literary heritage, which raised a number of fundamental questions. Which Attic authors should be imitated, and which authors should not be imitated? In what respects and to what extent should the various selected models be imitated? Cicero and Dionysius, whose treatises are the only extant texts from Late-Republican and Augustan Rome that explicitly address such questions, emphasize the tremendous variety within the category of classical Attic prose. In his Brutus , for example, Cicero provokingly asks: ‘Who are more unlike than Demosthenes and Lysias? Than either of them and Hyperides, than all of these and Aeschines? Whom then are you going to imitate? If one only, do you mean that all the others did not speak pure Attic? If all, how can you imitate them when they are so unlike each other?’ 10 Dionysius sets out to answer similar questions in On Imitation and On the Ancient Orators : ‘Who are the most

9 Cic. Orat. 23: Quo ne Athenas quidem ipsas magis credo fuisse Atticas . The translation is mine. 10 Cic. Brut. 285: Nam quid est tam dissimile quam Demosthenes et Lysias? Quam idem et Hyperides? Quam horum omnium Aeschines? Quam igitur imitaris? Si aliquem, ceteri ergo Attice non dicebant? Si omnes, qui potes, cum sint ipsi dissimillimi inter se? Cicero continues to ask such questions up to Brut. 291.

HERO VERSUS ZERO important orators and historians? What manner of life and style of writing did they adopt? Which characteristics of each should we imitate, and which should we avoid?’ 11 If the critics and rhetoricians in Rome could draw on such a large pool of Attic prose models, it seems all the more surprising that Demosthenes should emerge as such a dominant model. Sure enough, the reputation of Demosthenes as the king of eloquence was not unchallenged. Cicero, for instance, notes that Calvus and his movement of self-proclaimed Attic orators preferred the simple oratory of Lysias, while there were others on the Forum who imitated the obscurity of Thucydides or the sweetness of Xenophon. 12 Caecilius, Demetrius and Longinus do not profess a clear allegiance to a single model, but Demosthenes is certainly among their favorites. Caecilius, for instance, was not only a staunch defender of Lysias, but he also wrote several treatises on Demosthenes (section 1.5). Demetrius, in addition, presents Demosthenes as the principal exponent of the ‘forceful style’ ( χαρακτὴρ δεινός ), while Longinus numbers him among the main models of sublime literature. 13 All in all, the dominance of Demosthenes in the stylistic discourse is undeniable. 14 Even Calvus, whom Cicero presents as a passionate aficionado of Lysias, did not spurn Demosthenes: both the elder Seneca and the younger Pliny recognize in Calvus an imitator of Demosthenic vigor. 15

11 Dion. Hal. Ant. orat. 1.5: Τίνες εἰσὶν ἀξιολογώτατοι τῶν ἀρχαίων ῥητόρων τε καὶ συγγραφέων καὶ τίνες αὐτῶν ἐγένοντο προαιρέσεις τοῦ τε βίου καὶ τῶν λόγων καὶ τί παρ ’ ἑκάστου δεῖ λαμβάνειν ἢ φυλάττεσθαι . Dionysius lists the contents of Imit. at Pomp. 3.1: he discusses the nature of imitation (book 1), the authors that should be imitated (book 2), and the process of imitation (book 3). 12 On the self-styled Atticists see esp. section 1.4 above and sections 3.4, 4.5.3 and 5.6 below. For the Roman imitators of Thucydides and Xenophon, see Cic. Orat. 30–32. Thucydides influenced several Roman historians, such as Sallust and Quintus Aelius Tubero: see Leeman (1963) 179–187 and De Jonge (2017) 644–646. 13 Demetr. Eloc. 240–342 refers more often to Demosthenes than to any other author as an example of ‘forcefulness’ ( δεινότης ), although the critic does not associate his stylistic registers explicitly with individual authors, as Dionysius does. Demosthenes is Longinus’ main model for sublimity through emotion: cf. Innes (2002) 278. Long. Subl. 36.2 presents Demosthenes as a ‘flawed genius’ on a par with Plato and Homer. Wooten (1991) explains that Demetrius and Longinus particularly admire Demosthenes for his ‘conciseness’ ( συμτομία ) and ‘abruptness’ ( διάλυσις ). Innes (2002) argues that Demosthenes is Caecilius’ main model (cf. section 1.5 above), while Longinus has a more varied taste. See also Porter (2016) 105–106 on Longinus’ preference for Demosthenes as a rhetorical model. 14 The general attitude toward Demosthenes evolved from ambivalent in the Hellenistic era to laudatory in Late- Republican Rome: this has been the communis opinio since Drerup (1923). Cf. section 2.1 n. 1 above. 15 Sen. Controv. 7.4.8, Plin. Ep. 1.2.2: see section 5.2 below. The references to Demosthenes show that Calvus’ Atticism is not as monolithic as Cicero wants his audience to believe it to be: indeed, the concept of ‘Attic’ style

CHAPTER TWO

To understand what sets Demosthenes apart from the other Attic orators, we should turn to Cicero and Dionysius: it is in their works, after all, that Demosthenes is set up as the ultimate paragon of Athenian genius. In 46 BC, when Cicero was preoccupied with his polemic against the so-called Atticists, he emphatically equates the style of Demosthenes with the greatness of Athens, claiming that his opponents could benefit from reading Demosthenes’ speeches: ‘Let them learn what is Attic, and measure eloquence by his strength, not by their own weakness.’ 16 Moreover, Cicero presents his translation of Demosthenes’ speech On the Crown as a ‘norm by which to measure the speeches of those who may wish to speak in the Attic manner’. 17 In a flash of sarcasm, Cicero expresses his disbelief at the Atticists’ neglect of Demosthenes as a model for their oratory: ‘As if Demosthenes comes from Tralles!’ 18 Needless to say, the famous orator was not an inhabitant of this Lydian town, but a prominent citizen of Athens. Indeed, according to Cicero and Dionysius, he is the man who embodies the city’s legacy more than any other Attic author, admiring his oratorical genius, but also the political and cultural values that he stands for, i.e., democracy and freedom (section 5.5) Faced with Caesar’s dictatorship and its aftermath, Cicero turned to Demosthenes as a model for his own role in the city. It is not for nothing that Cicero calls his fourteen speeches against Marc Antony (44–43 BC) his Philippics , after Demosthenes’ orations (351 BC) against the expansive politics of king Philip of Macedon: in his Latin Philippics , Cicero presents himself as a champion of Roman freedom, just as Demosthenes defended the independence of Athens in his original Greek Philippics. 19 Dionysius, too, pays attention to Demosthenes’ role as a guardian of Athens. The samples that he quotes from Demosthenes’

is a flexible construct that can incorporate diverse elements from the Attic literary repertoire. On the ‘vigor’ ( vis ) of Demosthenes, see sections 2.3.2 and 2.3.3 below. 16 Cic. Orat. 23: Quid enim sit Atticum discant eloquentiamque ipsius viribus, non imbecillitate sua metiantur. 17 Cic. Opt. gen. 23: Erit regula, ad quam eorum dirigantur orationes qui Attice volent dicere. 18 Cic. Orat. 234: Quasi vero Trallianus fuerit Demosthenes. For the use of geographic names as shorthand formulas for Asian rhetoric, see section 1.2 above on Cic. Orat. 25 and Dion. Hal. Ant. orat. 1.7 (Mysia, Phrygia and Caria). 19 Cicero refers to the speeches as Philippics in Ad Brut. 2.4.2. They were also known as his Speeches against Antony (Orationes in Antonium or Antonianae ): e.g. Sen. Suas. 6.15, Gell. NA 1.22.17. On Demosthenes as a political model for Cic. Phil. , see Wooten (1983) and Ramsey (2003) 16–18. Cf. Wooten (1977), Dugan (2005) 309–314, Bishop (2016) and section 5.5.1 below on Cicero’s use of Demosthenes as a model to negotiate his own role in Caesar’s Rome (46 BC). Usher (2008) argues that not Dem. Phil. , but rather Dem. De cor. was by far the most influential of Demosthenes’ speeches for Cic. Phil. : it seems that Dem. De cor. was the text Cicero admired on two accounts, both as a political and as a stylistic masterpiece (cf. Cic. Orat. 133).

HERO VERSUS ZERO speeches often have a notable patriotic flavor. 20 From Demosthenes’ Third Olynthiac (349 BC), for instance, Dionysius quotes a passionate appeal to the Athenians to take up arms against Philip, noting: ‘If we, who are so far removed in time and unaffected by the events, are so carried away and overpowered that we follow wherever the speech leads us, how must the Athenians and the rest of the Greeks have been excited at the time by the orator addressing them on live and personal issues.’ 21 Dionysius and Cicero clearly not only admire Demosthenes for his style, but also for his politics. 22 Indeed, Demosthenes could be seen as the last true exponent of Athenian democracy and freedom: after all, the conquests of Philip and his son Alexander brought an end to the independence of the Greek poleis, and ushered in the rule of Eastern monarchs over the Greek world. This is where Hegesias comes into the picture: a generation younger than Demosthenes, his floruit roughly coincides with the alleged beginning of literary decline from the Classical into the Hellenistic era. Thus, according to the authors of the first centuries BC and AD, Hegesias was a founding father of the rhetorical and cultural decline of Asianism. Strabo (64/63 BC–ca. 24 AD) paints the default picture of Hegesias as ‘the orator who more than any other initiated the so-called Asian style ( Ἀσιανὸς ζῆλος ), whereby he corrupted the established Attic custom ( τὸ καθεστὼς ἔθος τὸ Ἀττικόν)’. 23 In other sources, too, Hegesias is

20 Cf. section 5.5.2 below. Like Cicero, Dionysius was fond of De cor. : see esp. Dion. Hal. Dem. 14.1. He quotes De cor. 1 three times (at Comp. 6.9, 18.17–22 and 25.26): ‘First of all, men of Athens, I pray to all the gods and all the goddesses …’ ( πρῶτον μέν , ὦ ἄνδρες Ἀθηναῖοι , τοῖς θεοῖς εὔχομαι πᾶσι καὶ πάσαις …). The passage is also cited by Quint. Inst. orat. 9.4.63. De Jonge (2008) 340–347 shows that Dionysius gives two incompatible scansions of this passage, demonstrating that the local contexts can account for the differences: cf. section 2.4.3. See also Long. Subl. 17–18 for patriotic passages from Dem. Phil. I and esp. De cor. 208 (the oath by the fighters of Marathon): Whitmarsh (2001) 57–71 connects the politics of Demosthenes to Longinus’ conception of the sublime. Yunis (2019) shows that Dionysius uses Demosthenes as a model ‘not only of wonderful, effective writing, but also of what the citizens of the Augustan empire, both Greeks and Romans, can achieve.’ 21 Dion. Hal. Dem. 22.4: Ὅπου γὰρ ἡμεῖς οἱ τοσοῦτον ἀπηρτημένοι τοῖς χρόνοις καὶ οὐθὲν πρὸς τὰ πράγματα πεπονθότες οὕτως ὑπαγόμεθα καὶ κρατούμεθα καὶ, ὅποι ποτ ’ ἂν ἡμᾶς ὁ λόγος ἄγῃ, πορευόμεθα , πῶς τότε Ἀθηναῖοί τε καὶ οἱ ἄλλοι Ἕλληνες ἤγοντο ὑπὸ τοῦ ἀνδρὸς ἐπὶ τῶν ἀληθινῶν τε καὶ ἰδίων ἀγώνων . The passage that Dionysius quotes is Dem. Olynth. III 23–32. 22 Cf. sections 5.5.1 and 5.5.2 below on the different ways in which Cicero (under Caesar) and Dionysius (under Augustus) interpret the political legacy of Demosthenes. 23 Str. Geogr. 14.1.41: Ἡγησίας τε ὁ ῥήτωρ , ὅς ἦρξε μάλιστα τοῦ Ἀσιανοῦ λεγομένου ζήλου παραφθείρας τὸ καθεστὼς ἔθος τὸ Ἀττικόν . Strabo was active in Rome at least between 44 and 29 BC: for (the dates of) his live, see Dueck (2017) 1. Strabo’s statement that Hegesias was native to Magnesia on the river Meander (Str. Geogr. 14.1.61 = T1 Jacoby) must be mistaken: Hegesias himself tells us that he came from Magnesia on the river

CHAPTER TWO depicted as a leading Asianist: Aelius Theon (probably first century AD), for instance, refers to ‘the orator Hegesias and the so-called Asian orators’ ( Ἡγησίας ὁ ῥήτωρ καὶ οἱ Ἀσιανοὶ καλούμενοι ῥήτορες ). 24 To be sure, Hegesias is not the only author from the Hellenistic era, who is severely criticized in later times: Philodemus, Dionysius and Longinus, for instance, supply the names of several other Hellenistic orators and historians whose styles they consider below par. 25 Yet, no author is as fiercely criticized and as confidently equated with Asian degeneracy as Hegesias. Typically, Cicero associates six Greek authors with Asianism, but the only one he unequivocally rejects is Hegesias. 26 We should remember that the designation ‘Asian’ was mostly used as a derogatory term: ancient authors rarely referred to themselves as Asianists.27 Hegesias is no exception to this rule. On the contrary, Cicero pictures him as an aspiring Atticist, an imitator of the Attic orators Lysias and Charisius. Jestingly, Cicero suggests that Hegesias ‘regards himself so thoroughly Attic that he considered the native Attic writers almost uncouth rustics in comparison with himself’. 28 Cicero uses the example of Hegesias as a stern warning to the

Sipylus (Dion. Hal. Comp. 4.11 = F19 Jacoby). The word ζῆλος (LSJ ‘jealousy, rivalry, emulation, zeal’) is also associated with Asianism at Plut. Ant. 2.5. Cf. the title of Caecilius’ treatise How the Attic and Asian Styles Differ (Suda κ 1165 = Caecilius T1 Woerther), where the word ζῆλος is applied to both Attic and Asian style. For the various uses of ζῆλος in ancient literary theory, see Schippers (2019) 36–40, 45–50. 24 Theon Prog. p. 71.7–11 Spengel. On the date of Theon, see Kennedy (2003) 1: the terminus post quem is the late first century BC (Theon mentions Dion. Hal. and Theodorus of Gadara), 95 AD is a possible terminus ante quem, as a certain Theon, who wrote on stasis theory, is mentioned at Quint. Inst. orat. 3.6.48 and 9.3.46. The appearance of Hegesias in Theon’s text is connected to prose rhythm: see section 2.4.2 below. 25 Philod. Rhet. 4 col. 21.15–25 p. 180 Sudhaus lists Hegesias and the Hellenistic historian Clitarchus of Alexandria (next to the fourth-century rhetorician Alcidamas) as lovers of metaphor; Dion. Hal. Comp. 4.14–22 censures the lack of attention to word arrangement in several Hellenistic historians (Phylarchus, Duris, Polybius, Psaon, Demetrius of Callatis, Hieronymus, Antigonus, Heraclides, Hegesianax) and the Stoic philosopher Chrysippus: see Aujac and Lebel (1981) 203–204. Long. Subl. 3.2 charges Hegesias as well as the aforementioned Clitarchus, Callisthenes of Olynthus, Amphicrates of Athens and Matris of Thebes with the stylistic vice ‘tumidity’ ( ὄγκος ): cf. Mazzucchi (2010) 150–151 and section 5.6.1 below. 26 Cic. Brut. 325 mentions Timaeus, the brothers Hierocles and Menecles of Alabanda, Aeschylus of Cnidus and Aeschines of Miletus. On Cicero’s discussion of the two ‘types of Asian styles’ ( genera Asiatica dictionis ) and his ambivalent verdict on the aforementioned ‘Asian’ authors: cf. section 1.6 above. For the Roman authors, who are associated with Asianism, such as Hortensius, Marc Antony and Maecenas, see section 5.2 below. 27 The only recorded example is the Augustan declaimer Craton, who according to Sen. Controv. 10.21 was a ‘professed Asianist’ ( professus Asianus ). Cf. section 5.2 n. 11 below. 28 Cic. Brut. 286: Isque se ita putat Atticum, ut veros illos prae se paene agrestis putet. According to Cicero, Hegesias ‘wants to imitate Lysias’ ( imitari Lysiam volt ) and he ‘wants to be like Charisius’ ( Charisi vult esse

HERO VERSUS ZERO self-styled Atticists in Rome: even those who fashion themselves genuine Attic orators, may in fact be poorly disguised Asianists. Cicero’s mockery of Hegesias’ Atticist aspirations makes it abundantly clear that the categories of Attic and Asian oratory are highly fluid: for Cicero, however, it is self-evident that the man who stood at the cradle of the Hellenistic kingdoms and the cultural dominance of Asia is an Asianist. 29 Cicero connects the style of Hegesias to several cultural stereotypes about Asia, such as servitude, greed and insolence: he describes Hegesias and his fellow-Asianists, for instance, as ‘slaves to rhythm’ ( numero servientes ), while he associates him elsewhere with desire for gold ( aurum ). The fact that the man from Magnesia considered himself more Attic than the native Attic orators, lastly, fits with the bias that the peoples of Asia were generally rude, vain and hubristic. 30 Dionysius does not explicitly refer to Hegesias as an Asian author, but he does implicitly associate him with Asia. The critic consistently connects the style of Hegesias to immoral behavior, describing the ‘Hegesian style’ (Ἡγησιακὸν σχῆμα ) of word arrangement, for instance, as ‘precious’ ( μικρόκομψον ), ‘ignoble’ (ἀγεννής ) and ‘effeminate’ ( μαλθακόν ): these words resonate with Dionysius’ description of Asian rhetoric in the preface of On the Ancient Orators (section 1.6). 31 In addition, we have seen that Dionysius claims that Hegesias’ words ‘are likely to be uttered only by women or emasculated men’. 32 Dionysius also quotes one extensive passage from Hegesias’ history of Alexander. 33 This passage has a similis ), who was himself an imitator of Lysias: Brut. 286, Orat. 226. Wooten (1975) 95–98 suggests that Hegesias’ style is indeed a continuation of the style of Charisius (and Cleochares), but there is too little surviving material from Charisius to substantiate this claim. 29 Further evidence of Hegesias’ interest in Athens can be found in the titles of the dialogue The Lovers of Athens, Aspasia and Alcibiades , attributed to him in a Rhodian inscription containing a catalogue of various writings (T11 Jacoby). Staab (2004) thinks that the majority of fragments of Hegesias (esp. F6–13 Jacoby) come from The Lovers of Athens , arguing that the work mocked Demades, Demosthenes and other important characters of Athenian politics in the second half of the fouth century BC: thus, Hegesias’ affected style (cf. section 2.4 below) may have been an intentional parody. Staab’s argument, which is built on a study of the word φιλαθήναιος in Greek poetry and prose of the fourth century BC, does not immediately replace the traditional view that most of Hegesias’ fragments come from his historiographical work, but Staab does demonstrate how little we know about the nature of Hegesias’ work other than the prejudice that we find in his later detractors. 30 See Cic. Orat. 230 (on slavery) and Att. 12.6.1 (on gold): cf. section 2.4.2 below. For the Asian stereotypes in Greek and Roman thought, see section 5.3 below. 31 Dion. Hal. Comp. 4.12. For Dionysius’ view of Hegesias’ style of word arrangement, see section 2.4.3 below. 32 Dion. Hal. Comp. 18.28: Ὑπὸ γυναικῶν ἢ κατεαγότων ἀνθρώπων λέγοιτ ’ ἂν. 33 Dion. Hal. Comp. 18.26 (= F5 Jacoby). Dionysius compares the rhythm of Hegesias to Hom. Il. 22.395–411: in both passages, a victor (Alexander the Great resp. Achilles) drags the body of his vanquished foe (the eastern

CHAPTER TWO particularly Asian color: it relates how Alexander decided to tie an Eastern king with his feet to a chariot and to drag him around on his encampment. Hegesias describes how the king ‘was yelling like a barbarian, begging Alexander for mercy and addressing him as lord’ ( βάρβαρον δ’ἐβόα , δεσπότην καθικετεύων ). Thus, Dionysius gives his readers a sample of slavish behavior, which was considered quintessentially un-Greek. Hegesias goes on to describe the king’s ‘strange language’ ( σολοικισμός ) and ‘his fat and the hollow span of his belly’ ( τὸ στέαρ καὶ τὸ κύτος τῆς γαστρός ), which makes the man resemble a ‘Babylonian beast rather than a man’ (βαβυλώνιον ζῷον ἕτερον ἀνδρός ). By quoting this particular passage, Dionysius reinforces the suggestion that Hegesias and his works are extravagant, unmanly and undignified—in a word, deeply Asian. To conclude, Demosthenes and Hegesias owe their prominence in the stylistic discourse of Late-Republican and Augustan Rome to a large extent to their positions on either side of the putative watershed between the Classical and Hellenistic eras: Dionysius was considered the last true champion of Athenian freedom, whereas Hegesias was firmly associated with the subsequent rule of the Eastern monarchs over the Greek world. The surviving judgments of their styles evoke the clichés of Athens and Asia (section 5.3), connecting Demosthenes to such virtues as freedom, democracy and good taste, and Hegesias to such vices as slavery, monarchy and impertinence. These stock features are also highlighted in the quotations from their works: the passages are not only selected on the basis of their stylistic attributes, but also because their contents call to mind the Attic and Asian values with which their authors are identified. Yet, Cicero and Dionysius do not only build their judgment on cultural bias, but also on stylistic analysis: what do they appreciate about the style of Demosthenes, and what do they disparage in the style of Hegesias?

2.3 How To Compose Great Prose: Demosthenes as the Canon of Style According to Quintilian, Athens may have produced ten great orators, but Demosthenes is ‘by far the greatest’ ( longe princeps ): he is the ‘standard of oratory’ ( lex orandi ). 34 In this section,

king resp. Hector) behind his chariot, but according to Dionysius the rhythm in Hegesias displays ‘baseness’ (ταπεινότης ), whereas Homer’s lines ooze ‘nobility’ ( εὐγενεία ). Cf. sections 2.4.2 and 2.4.3 below. 34 Quint. Inst. orat. 10.1.76. There is much disagreement about the role of Caecilius of Caleacte in the creation of the canon of ten Attic orators. Esp. the precise nature of the work On the Stylistic Character of the Ten Attic Orators , attributed to him in Suda κ 1165 (= T1 Woerther) is a bone of contention: the debate is succinctly summarized by Woerther (2015) 46–47 n. 8. While Douglas (1956) and Smith (1995) attribute the origins of the canon to different periods (Douglas second century AD, Smith third/second century BC), several scholars have

HERO VERSUS ZERO

I will turn to Cicero’s and Dionysius’ verdicts on Demosthenes’ style, which, as we will see, are consistent with Quintilian’s: Cicero and Dionysius consider the works of Demosthenes a veritable canon of style. 35 Thus, the variety that was thought to characterize the literary legacy of Classical Athens could be found in the works of a single man. The aim of this section is not to present an exhaustive list of the stylistic criticisms of Demosthenes in the works of Cicero and Dionysius; rather, I will focus on the most prominent points that the critics raise, including his almost kaleidoscopic versatility, his forcefulness in high-stakes contexts, and his knack for word arrangement. After a brief survey of the Hellenistic reception of Demosthenes’ oratory (section 2.3.1), I will successively explore the discussions of Demosthenes’ style in the rhetorical treatises of Cicero (section 2.3.2) and in the critical essays of Dionysius (section 2.3.3).

2.3.1 Demosthenes in Hellenistic Scholarship The critical reception of Demosthenes’ speeches starts in his own time. In general, his eloquence seems to have impressed his contemporaries, including his rivals and opponents: although such men as Aeschines, Dinarchus and Hyperides all have axes to grind with Demosthenes, they acknowledge that he is a ‘skillful’ speaker ( δεινός ). 36 In Against Ctesiphon , Aeschines describes him as ‘skillful of speech, infamous of life’ ( δεινὸς λέγειν , κακὸς βιῶναι ): Aeschines makes Demosthenes’ rhetorical skills suspicious by exposing the convincingly argued that Caecilius may have played an important role in the establishment of the canon: Worthington (1994) points out that the traces of the canon’s existence do not appear before the first century BC, and O’Sullivan (1997) adopts a nuanced approach, arguing that ‘the exact contribution that Caecilius made to this conception cannot now be known; he may well have been behind particular lists of writers recommended for imitation, but such lists were subject to variation and were only the expressions of deeper ideas’. 35 The one possible weakness of Demosthenes, according to Cic. Orat. 90, is his lack of humor: cf. Wooten (1997) 184–185. Cicero defends Demosthenes by claiming that the orator was not so much ‘witty’ ( dicax ) as he was ‘humorous’ ( facetus ): Cicero attributes the former to talent, and the latter to training. This judgment fits well with the general view that Demosthenes’ genius was a product of technique rather than of nature: cf. section 2.3.1 below. On his lack of humor, see also Long. Subl. 34.3, who also mentions his incompetence in the domain of characterization, his lack of fluency, his unsuitability for epideictic oratory and his lack of charm. 36 Anastassiou (1966) 55–66. See, e.g., the references to Demosthenes as a ‘skillful speaker’ ( δεινὸς λέγειν ) in Aeschin. Ctes. 174 and 215, ‘verbal trickster’ ( τεχνίτης λόγων ) in Aeschin. Tim. 170 and Ctes. 200, ‘outstanding speaker’ ( περιττὸς ἐν τοῖς λόγοις ) in Aeschin. Tim. 119 and Fals. leg. 114. Note that the word δεινός is not yet used with its later stylistic connotation of forcefulness (cf. esp. Demetrius, Hermogenes): see Voit (1934) 12–15, who connects the negative connotation of δεινός in the fourth century BC to the general mistrust of the art (or trickery) of the sophists. Cf. Grube (1961) 136–137.

CHAPTER TWO deceitful, dishonorable conduct that his words supposedly aim to cover up. 37 In the Hellenistic period, Aeschines’ cynical representation of Demosthenes landed on fertile ground: Demosthenes was, especially among Peripatetic philosophers, often considered a corrupt politician who seduced his audience with verbal trickery. 38 The scarcity of our Hellenistic sources makes it difficult to follow the development of Demosthenic criticism in the centuries after his death, but it seems that Demosthenes’ style remained suspect for a long time: criticized by such philosophers as Theophrastus and Demetrius of Phalerum, he seems to have received much acclaim among the rhetoricians and literary critics of the second century BC, only to emerge from the mist as an all-round genius in the first century BC. 39 There are several elements in the Hellenistic treatment of Demosthenes that seem to foreshadow the extreme praise that he receives in the works of Cicero and Dionysius. First, his ancient biographers, beginning with Demetrius of Phalerum, almost invariably mention that the orator had to overcome many natural defects in order to become a skilled speaker: he stuttered, he had a shortness of breath, he could not pronounce his rhos, he was unable to project his voice and he had a shoulder spasm, which forced him to subject himself to bizarre exercises in order to train his voice and body. 40 These stories are almost certainly romantic fictions, but they have to a large extent determined the later approaches to his style. Already

37 Aeschin. Ctes. 174. Cicero and Dionysius were familiar with Aeschines’ insulting depictions of Demosthenes: Cicero translates Aeschin. Ctes. (cf. Cic. Opt. gen. ), Dionysius quotes several passages from Aeschin. Ctes. to show that even Aeschines was aware of Demosthenes’ amazing ability as a speaker (Dion. Hal. Dem. 35). 38 See esp. Cooper (2000) 234–238. Lossau (1964), who attempts to reconstruct the early exegetical tradition of Demosthenes’ speeches, argues that the Hellenistic assessment of Demosthenes is not as negative as is often assumed, but Lossau’s reconstructions of Peripatetic and Alexandrian scholarship on the basis of later sources are unwarranted, as Gibson (2002) 31 shows. The latter offers an edition with translation and commentary of the papyri containing the exegesis of Arius Didymus (second half of the first century BC) and later commentators. 39 Anastassiou (1966) 66–81 discusses the criticism of Demosthenes in the early Peripatos. According to Cooper (2000) 238, Demosthenes owed his rehabilitation to the fact that his proficiency as an orator was the product of training rather than talent: ‘By the second century, Demosthenes would become the darling of rhetoricians, whose own reputations depended on successfully teaching rhetoric and on showing how the skill could be learned.’ The Peripatetic reproach of Demosthenes continues in Dionysius’ day, as is evidenced by Amm. I 1.1. 40 See Demetrius of Phalerum fr. 165 Wehrli, Libanius 296.62, Zosimus 299.60, Anonymous Vita 305.66, Suda δ 456, Plut. Mor. 844d–e. Cf. Dion. Hal. Dem. 53.4, who mentions as his sources ‘Demetrius of Phalerum and all other biographers’ ( Δημήτριός τε ὁ Φαληρεὺς καὶ οἱ ἄλλοι πάντες οἱ τὸν βίον αὐτοῦ συγγράψαντες ). See Pomeroy (1997) 173–174, who lists the stories about Demosthenes’ struggle with his physical defects, concluding that it is unlikely that ‘an adolescent would invent bizarre devices’, such as reciting with a mouthful of pebbles, discoursing while running, and hanging a sword over his shoulder to correct his shoulder spasm.

HERO VERSUS ZERO at an early stage he was characterized as ‘painstaking rather than naturally gifted’ ( ἐπιμελὴς μᾶλλον ἢ εὐφυής ): his outstanding oratorical performance was more the product of technique (τέχνη , ars ) than of any natural talent ( φύσις natura ). 41 Interestingly, Cicero and Dionysius continue to focus on Demosthenes’ application of rhetorical technique: according to them, Demosthenes had learnt every trick in the rhetorical handbook and his speeches can thus be adduced to illustrate virtually all aspects of stylistic theory. 42 Another feature that persisted throughout the history of ancient scholarship on Demosthenes is the custom of comparing him to other authors ( σύγκρισις ): we have seen that Caecilius compares him to Aeschines and to Cicero, while Dionysius compares samples of his works to thematically related passages from Lysias, Thucydides, Isocrates and Plato. 43 In the third century BC, Demosthenes was most often compared to Isocrates: Cleochares of Myrlea is credited with one of the earliest explicit comparisons, claiming ‘that the speeches of Demosthenes are a lot like the bodies of soldiers, whereas the speeches of Isocrates resemble the bodies of athletes’. 44 The contrast between training for war and training for parades and festivals touches on the dichotomy between practical and epideictic oratory: Demosthenes is usually presented as a particularly effective exponent of the former genre, whereas Isocrates is considered to be more at home in ceremonial settings. 45 In the first century BC, Cicero and

41 Suda δ 455. Cf. Cooper (2000) 241 n. 4: the extant biographies usually draw a clear contrast between his ‘nature’ ( φύσις ) and his ‘exercise’ ( μελέτη ). 42 We have seen that Cic. Orat. 90 uses the distinction between nature and training to explain Demosthenes’ alleged lack of humor: section 2.3 n. 35 above. Dion. Hal. Amm. I aims to refute the claim of an unnamed Peripatetic that Demosthenes’ eloquence is based on the orator’s knowledge of Arist. Rh. , but at Amm. I 3.1, he does submit that Demosthenes’ speeches are indebted to ‘other introductory treatises’ ( ἑτέραι τινὲς εἰσαγωγαί ). 43 Οn the importance of σύγκρισις as a critical tool, see Focke (1923) and Vardi (1996). See also De Jonge (2018c) on σύγκρισις as a form of competition between Greece and Rome. Comparison was a favorite tool of both Dionysius (cf. Pomp. 1.9) and Caecilius ( Suda κ 1165 = T1 Woerther): see section 1.5 above. The comparisons between Demosthenes and other Attic prose authors occupy a substantial part of Dion. Hal. Dem .: see Dem. 9–10 (with Thucydides), 11–13 (with Lysias), 17–22 (with Isocrates) and 23–32 (with Plato). 44 Phot. Bibl. cod. 176: ( Κλεοχάρης δὲ ὁ Σμυρλεανός φησι ) τοὺς μὲν Δημοσθενικοὺς λόγους τοῖς τῶν στρατιωτῶν ἐοικέναι μάλιστα σώμασι , τοὺς Ἰσοκρατικοὺς τοῖς τῶν ἀθλητῶν. The comparison is attributed to Philip of Macedon in [Plut.] X orat. 845d and may be related to the distinction between ‘real bodies’ ( ἀληθινὰ σώματα ) and ‘statues’ ( ἀδριάντες ) in Alcidamas (fourth century BC): see O’Sullivan (1992) 79. The comparison between military and ceremonial bodies/weapons became a commonplace in stylistic theory, cf. Wollner (1886): e.g. Cic. De or. 2.94, Βrut. 37, Orat. 42, Dion. Hal. Dem. 32.1, Quint. Inst. orat. 10.1.33. Cf. section 2.3.3 below. 45 Cf. Yunis (2019) 101–104. On the distinction between the practical, or ‘performative style’ ( λέξις ἀγωνιστική ) and the epideictic, or ‘written style’ ( λέξις γραφική ) in Arist. Rh. 3.12, see Innes (2007) 151–156.

CHAPTER TWO

Dionysius also discuss Demosthenes’ aptitude as a model for real-life speeches on urgent matters: it is on this kind of actively involving, performative oratory that both Cicero and Dionysius focus. 46 Already during his own lifetime, Demosthenes was regarded, as we have seen, as ‘skillful’ ( δεινός ); in the centuries after his death, the word δεινός remained inseparably connected to his name, but its meaning gradually shifted to denote ‘vigor’ or ‘forcefulness’ (δεινότης , vis ). 47 The idea that Demosthenes has the power to overwhelm his audience is clearly articulated in Demetrius’ On Style , which presents the orator as the principal model of the ‘forceful style’ ( χαρακτὴρ δεινός ), whereas he is rarely mentioned in connection to the other three stylistic registers. 48 We will presently see that Cicero and Dionysius also stress the tremendous power of Demosthenes’ style, highlighting the spellbinding lightning-like quality of his oratory. After Cicero and Dionysius, the identification of Demosthenes with forcefulness remained a commonplace feature of literary criticism: in his work On Types of Style , Hermogenes of Tarsus (second century AD) presents forcefulness ( δεινότης ) as the style that constitutes an ideal combination of all stylistic types, most perfectly exemplified by Demosthenes. 49

2.3.2 Cicero on Demosthenes’ Style Cicero’s discussions of Demosthenes’ style can mainly be found, as we saw, in his three rhetorical treatises of 46 BC (section 1.4). In these works, there are two major stylistic features that Cicero associates with Demosthenes: he emphasizes the orator’s versatility and

46 See Ooms and De Jonge (2013) 100–102 for Dionysius’ focus on ‘speeches suited to performance’ ( ἐναγώνιοι λόγοι ). For the similar focus in Cicero on ‘real-life forensic cases’ ( verae causae forenses ), see Cic. Orat. 12–13, 30, 32, 36, 62, 120, 143, 148, 170, 208–209, 221. Cf. esp. sections 3.6 and 5.5.1 below. 47 For the evolution of the meaning of δεινός and its cognates, see esp. Voit (1934), who argues that the word originally carried a meaning similar to that of δυνατός (‘capable’) and was a virtual antonym of ἰδιώτης (‘layman’): according to Voit, δεινός appears as a technical term for the first time in Demetr. Eloc. (which he mistakenly attributes to Demetrius of Phalerum, fl. ca. 300 BC), specifically denoting the agonistic, passionate style of Demosthenes. 48 Demetr. Eloc. 240–342. Cf. section 2.2 n. 13 above. 49 Rutherford (1998) connects Hermogenes’ discussion of Demosthenes to the contemporary literary interests of the Second Sophistic: Demosthenes’ proficiency in all stylistic types, for instance, can be linked to the focus on ‘versatility’ (ποικιλία ) in the Antonine era. Rutherford also discusses the surviving texts of [Aristid.], who favored Demosthenes as well as Plato. For the all-encompassing nature of the forceful type of style, see the convenient diagram of Wooten (1987) xii.

HERO VERSUS ZERO his tremendous force, which is connected to his talent for word arrangement. The former element touches on Cicero’s very definition of the ideal orator: ‘He will be eloquent, who can discuss commonplace matters simply, lofty subjects impressively, and topics ranging between in a tempered style.’ 50 Thus, Cicero’s consummate orator masters all three oratorical styles (genera dicendi ), for which Cicero uses a versatile terminology (section 3.2): they are now commonly referred to as the ‘plain’ (e.g. genus subtile ), the ‘grand’ (e.g. genus grave ) and the ‘intermediate style’ (e.g. genus mediocre ). In chapter three, we will have a closer look at Cicero’s application of this three-style doctrine (esp. section 3.4), but for now, we should simply note that Cicero’s threefold scheme closely connects style to substance: the language of the orator should be adapted to the subject matter at hand. 51 An orator who masters all three types of style, will be able to speak in an appropriate manner about any topic on any occasion. 52 According to Cicero, certain men have been successful in individual registers, but it is virtually impossible to attain perfection in all three types: as we will see, Cicero asserts that Lysias triumphed only in the simple style, Pericles merely in the grand style and Demetrius of Phalerum only in the intermediate style (section 3.3). Cicero insists that all-round perfection is something divine, a Platonic Form that we can only conceive in our minds, but that will never actually exist on the Forum: ‘As there is something perfect and surpassing in the case of sculpture and painting—an intellectual ideal by reference to which the artist represents those objects which do no themselves appear to the eye, so with our minds we conceive the ideal of perfect eloquence, but with our ears we catch only the copy.’ 53 Thus, no orator, dead or alive,

50 Cic. Orat. 100: Is est enim eloquens qui et humilia subtiliter et alta graviter et mediocria temperate potest dicere. Cf. Orat. 101 for a definition in similar words: Is erit igitur eloquens, ut idem illud iteremus, qui poterit parva summisse, modiaca temperate, magna graviter dicere (‘he, then, will be eloquent, to repeat my former definition, who can discuss trivial matters in a plain style, matters of moderate significance in the tempered style, and weighty affairs in the grand manner’). 51 Cf. Dion. Hal. Isoc. 12.4: Βούλεται ἡ φύσις τοῖς νοήμασιν ἕπεσθαι τὴν λέξιν , οὐ τῇ λέξει τὰ νοήματα (‘nature demands that the words should follow the thought, not that the thought should follow the words’). 52 Cf. section 3.2, table 3 below, where we will see that Cicero connects the three styles not only to three types of subject matter, but also to the so-called ‘functions of the orator’ ( officia oratoris ). On ‘appropriateness’ ( τὸ πρέπον , decorum ) as a guiding principle, see Cic. Orat. 70–74. 53 Cic. Orat. 9: Ut igitur in formis et figuris est aliquid perfectum et excellens, cuius ad cogitatam speciem imitando referuntur ea quae sub oculos ipsa non cadunt, sic perfectae eloquentiae speciem animo videmus, effigiem auribus quaerimus. Dell’Innocenti Pierini (1979) explains that Cicero’s reference to Plato’s theory of Forms is relevant to his discussion of the three stylistic types, as the Greek and Latin terms for Platonic Forms

CHAPTER TWO can wholly comply with Cicero’s high standards, not even Demosthenes: ‘Although he stands pre-eminent among all in every single style of oratory ( in omni genere dicendi ), he still does not always satisfy my ears, so greedy and insatiate are they and so often do they yearn for something vast and boundless.’ 54 Although even the best orator on earth can only create an imperfect reproduction of the perfect Form of oratory, Cicero finds that Demosthenes comes closest to this heavenly ideal. The rhetorician finds a veritable catalogue of samples from each of the three stylistic types in the orator’s oeuvre: ‘Many of his speeches are simple throughout—the one Against Leptines is an example; many are elevated throughout, as certain of his Philippics ; many are varied—the one against Aeschines On the False Embassy , and the one against Aeschines For Ctesiphon. The middle style he adopts whenever he will, and after an elevated passage he glides generally into this style.’ 55 Cicero singles out For Ctesiphon (now standardly known as On the Crown), as a tour de force of the three styles: ‘In his masterpiece For Ctesiphon , he began rather calmly, then in his discussion of the laws he continued rather concisely; after that, he gradually aroused the jury, and when he saw them on fire, throughout the rest of the oration, he boldly overleapt all bounds.’ 56 Although perfection is only theoretically possible, Cicero submits that orators can approach the ideal through an appropriate application of all three styles, and in this respect Demosthenes is second to none. 57 Cicero, then, praises Demosthenes because he is versatile in employing all three styles. The rhetorician reserves his highest admiration, however, for Demosthenes’ command of the grand style, in which his oratory has the power of thunderbolts ( fulmina ). Cicero warns any orator who employs the grand style to temper his impressive rhetoric with elements from the

(ἰδέαι , figurae ) are both used in stylistic theory to describe types of style: cf. Cic. Orat. 36, where the word forma is used as a translation of χαρακτήρ . On Cicero’s ideal orator as a Form, see Long (1995). 54 Cic. Orat. 104: Qui quanquam unus eminet inter omnis in omni gener dicendi, tamen non semper implet auris meas; ita sunt avidae et capaces et saepe aliquid immensum infinitumque desiderant. 55 Cic. Orat. 111: Multa sunt eius totae orationes subtiles, ut contra Leptinem; multa totae graves, ut quaedam Philippicae; multae variae, ut contra Aeschinem falsae legationis, ut contra eundem pro causa Ctesiphontis. Iam illud medium quotiens vult arripit et a gravissimo discedens eo potissimum delabitur. 56 Cic. Orat. 26: In illa pro Ctesiphonte oratione longe optima summissius a primo, deinde, dum de legibus disputat, pressius, post sensim incendens iudices, ut vidit ardentis, in reliquis exsultavit audacius. 57 Cf. Wooten (1977) 39: Cicero ‘discovered, I think for the first time, not only in Cicero’s life but in Demosthenic criticism, what is truly great about his (i.e., Demosthenes’) oratory’, that is, his versatility. Wooten (1997), however, describes Cicero’s treatment of Demosthenes as ‘tendentious and a defense of his own rhetorical practices’.

HERO VERSUS ZERO simple and intermediate styles, but still, it is in the highest register that rhetoric has the ‘greatest force’ ( maxima vis ), specifically the force to ‘sway’ (flectere ) the minds of the audience. 58 To Cicero’s mind, ‘force’ ( vis ) is precisely the stylistic virtue that is most typical for Demosthenes’ oratory: in this respect, Cicero agrees with the dominant trend in Demosthenic criticism that presents the orator as ‘forceful’ ( δεινός ), as can be seen in the Greek critics Demetrius, Dionysius and Hermogenes (section 2.3.1). This forcefulness can be compared to a ‘thunderbolt’ ( κεραυνός , fulmen ). 59 Cicero uses the image of lightning to describe the devastating effect of the grand style in general, and of Demosthenes’ forceful rhetoric in particular. In a letter to Atticus, Cicero presents Demosthenes’ thunderbolts as quintessentially Attic: ‘You have only to call to mind Demosthenes’ thunderbolts (Δημοσθένους fulmina ) to realize that a speaker can both be impeccably Attic ( Ἀττικώτατα ) and profoundly impressive.’ 60 The grand orator, wielding lightning bolts like a Zeus or Jupiter, can wreak havoc with his opponents and his audience. According to Cicero, therefore, the topics of the grand style are especially at home in real-life oratory, when the stakes are high. He adduces one of his own speeches as an example: ‘If they think that at the trial of Milo, when the army was stationed in the Forum and in all the temples round about, it was fitting to defend him in the same style that we would use in pleading a private case before a single referee, they measure

58 Cic. Orat. 69, 99. On the correspondence between Cicero’s doctrine of three styles and his doctrine of three oratorical tasks ( officia oratoris ), see section 3.2.1 table 3 below. On the dangers and the rewards that Cicero connects to the grand style, see esp. section 3.4 below. 59 See Porter (2016) 385–386: ‘The sublime attributes of Pericles, his capacity to thunder and send forth lightning with his magnificent oratory, were gradually assimilated to Demosthenes in the rhetorical tradition (possibly because Pericles’ speeches did not survive).’ For the association of the thunderbolt with Pericles, see e.g. Plin. Ep. 1.20.17–19 (= Eupolis fr. 94), Aristoph. Ach. 530–531, Plut. Per. 8.3–4 (= com. adesp. 701 Kassel- Austin), Cic. Brut. 44, Orat. 29, Quint. Inst. Orat. 12.10.65. Beside Cicero, the association of lightning with Demosthenes can be found in Longinus. At Subl. 12.4–5, the author compares Demosthenes to Cicero, claiming that the former ‘may be compared to a flash of lightning or a thunderbolt’ ( σκηπτῷ τινι παρεικάζοιτ ’ ἂν ἢ κεραυνῷ), while Cicero is more ‘like a widespread conflagration’ ( ὡς ἀμφιλαφής τις ἐμπρησμός ). At Subl. 34.4, Demosthenes is said to ‘out-thunder, as it were, and outshines orators of every age’ ( ὡσπερεὶ καταβροντᾷ καὶ καταφθέγγει τοὺς ἀπ’ αἰῶνος ῥήτορας ). Cf. ibid. 1.4: ‘A well-timed flash of sublimity shatters everything like a bolt of lightning and reveals the full power of the speaker at a single stroke’ ( ὕψος δέ που καιρίως ἐξενεχθὲν τὰ τε πράγματα δίκην σκηπτοῦ πάντα διεφόρησε καὶ τὴν τοῦ ῥήτορος εὐθὺς ἀθρόαν ἐνεδείξατο ). 60 Cic. Att. 15.1a.2: Sed si recordabere Δημοσθένους fulmina, tum intelleges posse et Ἀττικώτατα et gravissime dici. Cf. Cic. Orat. 21 (on lightning as a weapon of the grand style) and ibid. 235 (on lightning as characteristic for Demosthenes).

CHAPTER TWO the power of eloquence ( vis eloquentiae ) by their own limited ability, not by the nature of the art.’ 61 The other two types of style would not suffice in such circumstances: Cicero thinks that the plain style of Lysias is suitable only for ‘small private cases’ ( parvarum rerum causulae ), while the middle style is more suited to the ‘parade-ground’ ( palaestra ) than to the ‘battlefield’ ( arma ). 62 Thus, Cicero insists that the forcefulness of the grand style is required in urgent life-and-death cases: in a time when Rome is ruled by the weapons of Caesar, Cicero will not settle for anything less than a thunderous Demosthenes. 63 How does Demosthenes succeed in making his words rumble like thunder and flash like lightning? Cicero does not go into much detail, but in his Orator , he offers us a beginning of an answer: ‘Those famous thunderbolts ( fulmina illa ) of his would not have sped with such vibrant power, if they had not been whirled onward by rhythms ( numeri ).’ 64 Cicero’s remark comes at the end of a lengthy passage about the theory of word arrangement ( collocatio verborum ), which not only discusses rhythm, but also the method of combining individual words ( compositio ) and the creation of balance ( concinnitas ) in periodic sentences through figures of speech: 65 I will discuss the principles and motivations underlying Cicero’s views of word arrangement in more detail later (chapter 4). Cicero mentions Demosthenes on several occasions in the course of his long exposé on the intricacies of arrangement: he associates him with the practice of avoiding hiatus in combining words and with the practice of using appropriate rhythmical cadences (so-called clausulae) to round off his sentences. 66

61 Cic. Opt. gen. 10: Sed si eodem modo putant exercitu in foro et in omnibus templis, quae circum forum sunt, conlocato dici pro Milone decuisse, ut si de re privata ad unum iudicem diceremus, vim eloquentiae sua facultate, non rei nature metiuntur. ‘They’ in this sentence refers to the so-called Atticists, who only accepted simple, Lysianic oratory as truly Attic: see section 1.4 above, and sections 3.4, 4.4 and 5.6.1 below. According to ancient tradition (e.g. Asc. Mil. 41.24–42.2c), Cicero was thrown off balance by the unfavorable circumstances at the trial of Milo: the extant Pro Milone was written after the trial. La Bua (2010) 36–37 argues that Cicero uses Opt. gen. to rehabilitate himself: ‘Conscious of the canonicity that his speeches had achieved, he could not let stand the stigma conferred by his inability to deliver a good oration under difficult circumstances.’ 62 Cic. Opt. gen. 9 (on the simple style), Brut. 37 (on the intermediate style): cf. section 3.2.2 and 3.4 below. See also Cleochares’ comparison between Isocrates and Demosthenes (Phot. Bibl. cod. 176): section 2.3.1 above. 63 Cf. Bishop (2016) 190. 64 Cic. Orat. 234: Cuius non tam vibrarent fulmina illa, nisi numeris contorta ferrentur. 65 Cic. Orat. 149–236, more specifically ibid. 149–162 on compositio , ibid. 162–167 on concinnitas , ibid. 168– 236 on numerus. For the terminology, see ibid. 201. Cf. also section 4.2 below on the terminological incoherence of Latin discussions of word arrangement. 66 On hiatus, see Cic. Orat. 152, who claims that Demosthenes ‘generally avoids’ ( magna ex parte fugit ) hiatus: this accords well with Cicero’s insistence that hiatus in Latin prose is vicious. See however section 2.3.3 below

HERO VERSUS ZERO

Furthermore, Demosthenes is mentioned together with Isocrates and Aeschines as prime examples of ‘well-knit rhythmical prose’ ( numerosa et apta oratio ). 67 Thus, the devastating force of Demosthenes’ lightning bolts is derived, to Cicero’s mind, from his talent for word arrangement in general and prose rhythm in particular. Indeed, Cicero argues that these aspects are crucial in approximating his ideal, the Platonic Form of the orator: ‘Our words can win the praise at which we aim only by being properly arranged, and, as it were, bound together into a neat structure of words.’ 68 In the next section, we will see that the discussions of Demosthenes’ style in Dionysius overlap to a large extent with Cicero’s Orator , be it that the Greek critic goes into more detail about the orator’s rhythmical word arrangement.

2.3.3 Dionysius on Demosthenes’ Style The focus on Demosthenes is typical for Dionysius’ so-called middle period (section 1.5): he discusses the orator’s style at length in On Demosthenes and in the roughly contemporary treatise On the Arrangement of Words. 69 Stylistically, his praise of Demosthenes is built on three main pillars, that we have also encountered in Cicero’s treatment of the orator: Dionysius praises Demosthenes’ stylistic versatility, the suitability of his forceful rhetoric for life-and-death situations, and his talent for word arrangement. The first of these elements is stressed throughout On Demosthenes and On the Arrangement of Words . According to Dionysius, Demosthenes had created a style that was a on Dionysius’ discussion of hiatus in Demosthenes. On clausulae, see Cic. Orat. 226, where Demosthenes and Hegesias appear side by side. Cf. also section 4.4 below on Cicero’s and Dionysius’ insistence on ending periods on appropriate rhythmical endings. 67 Cic. Orat. 235. 68 Cic. Orat. 140: Sed haec nisi collocata et quasi structa et nexa verbis ad eam laudem quam volumus aspirare non possunt. The demonstrative haec refers to the foregoing passage ( Orat. 134–139) on stylistic ‘ornaments’ (lumina ), viz., figures of speech and figures of thought. 69 The middle-period treatise Amm. I is also concerned with Demosthenes: the letter does not discuss the orator’s style, but it supplies chronological arguments for Dionysius’ thesis that Demosthenes owed nothing to the rhetorical precepts of Arist. Rh. On the relative chronology of Dionysius’ works, see section 1.5 above. In his early works, Dionysius does not profess his preference for Demosthenes, as the extant early works are mainly concerned with the older orators Lysias, Isocrates and Isaeus. The brief discussion of Demosthenes at Imit. 5.4, however, does foreshadow his later praise: he notes that the orator ‘combines charm with dignity’ ( μετὰ τοῦ σεμνοῦ τὴν χάριν ἔχων ), which praises the orator’s versatility (cf. the present section and section 3.2 below), and he notes that Demosthenes possesses the qualities through which ‘juries are most often swayed’ ( μάλιστα δικασταὶ κατέχονται ), stressing the ‘practical use’ ( τὸ συμφέρον ) of his oratory (cf. the present section).

CHAPTER TWO perfect mixture of the various styles of his predecessors, such as Lysias, Thucydides, Isocrates and Plato: ‘Having found that public oratory had gone through a variety of changes, Demosthenes found himself following in the footsteps of some illustrious men, but refused to make any single orator or any single style his model, for he considered every one to be incomplete and imperfect, but instead he selected the best and most useful elements from all of them, weaving them together to make a single, perfect style.’ In this way, Demosthenes devised ‘a style that is a mixture of every form’ (χαρακτὴρ ἐξ ἁπάσης μικτὸς ἰδέας ). 70 The orator’s method of combining all kinds of style not only makes him an ideal model (as he exhibits everything that is praiseworthy about Attic prose), but it also exemplifies the method of imitation that Dionysius recommends to his readers: in the same way as the critic thinks that an original style can be created by studying and selecting the best qualities from the best authors, so did Demosthenes select the most appropriate features from every style and every author in order to devise his own unique form of eloquence. 71 As we will see in chapter three, Dionysius uses two three-style divisions to demonstrate the superior flexibility of Demosthenes’ oratory (esp. sections 3.2 and 3.5), which I will discuss only briefly here. With respect to his three ‘types of word selection’ (χαρακτῆρες τῆς λέξεως ), Dionysius uses a diverse terminology, calling one level ‘plain’ and ‘thin’ (e.g. λέξις ἀφελής , ἰσχνή ), another ‘elevated’ and ‘extraordinary’ (e.g. λέξις ὑψηλή , περιττή ), and a third ‘mixed’ (e.g. λέξις μικτή ), incorporating the best elements from the other two styles. 72 He ties up this tripartite scheme to his system of stylistic virtues ( ἀρεταί ),

70 Dion. Hal. Dem. 8.2–4: Τοιαύτην δὴ καταλαβὼν τὴν πολιτικὴν λέξιν ὁ Δημοσθένης οὕτω κεκινημένην ποικίλως , καὶ τηλικούτοις ἐπεισελθὼν ἀνδράσιν ἑνὸς μὲν οὐθενὸς ἠξίωσε γενέσθαι ζηλωτὴς οὔτε χαρακτῆρος οὔτε ἀνδρός , ἡμιέργους τινὰς ἅπαντας οἰόμενος εἶναι καὶ ἀτελεῖς, ἐξ ἁπάντων δ’αὐτῶν ὅσα κράτιστα καὶ χρησιμώτατα ἦν ἐκλεγόμενος συνύφαινε καὶ μίαν ἐκ πολλῶν διάλεκτον ἀπετέλει . Dionysius compares Demosthenes to the mutable sea-god Proteus: see section 3.5 below. Cf. section 2.4.3 below for Dionysius’ comparison between the power of word arrangement and Athena’s ability to change the appearance of Odysseus. 71 Cf. Dion. Hal. Imit. 1.1–5, which survives through a later epitome: he tells the story of (1) an ugly farmer, who fathers a beautiful child by showing his wife pictures of several beautiful women before sleeping with her, and (2) the painter Zeuxis, who painted a naked Helen by selecting and reproducing the beautiful body parts of various local girls. The anecdotes serve to show what can be achieved, ‘if one adopts what seem to be the best features of each of the ancient writers’ ( ἐπὰν ζηλώσῃ τις τὸ παρ ’ ἑκάστῳ τῶν παλαιῶν βέλτιον εἶναι δοκοῦν). The stories have recently attracted much attention: Hunter (2009) 107–127 discusses Dionysius’ subscription to Platonic imagery, Wiater (2011) 77–84 focuses on the notion of re-enacting the classical past, Schippers (2019) 1–11 expounds the origin of the stories as well as the critic’s conceptions of imitation and emulation. 72 Section 3.2, table 2 below. For the three types of word arrangement, see briefly the present section, and esp. section 3.2 and 3.3, and chapter 4 below.

HERO VERSUS ZERO distinguishing between virtues that are ‘essential’ ( ἀναγκαῖαι ) and those that are ‘additional’ (ἐπίθετοι ): according to Dionysius, the plain style exhibits only the essential virtues (e.g., clarity, purity), the elevated style focuses on the additional virtues (e.g., emotion, grandeur), and the mixed style is an appropriate combination of both classes of virtues.73 Consequently, the stylistic middle stands at the center of Dionysius’ attention: it is in this well-blended style, which is characterized by such values as ‘balance’ ( συμμετρία ), ‘timing’ ( εὐκαιρία ), ‘diversity’ (ποικιλία ) and ‘appropriateness’ ( τὸ πρέπον ), that he considers Demosthenes at his best.74 Thus, whereas Cicero had stressed the orator’s talent for grand oratory, Dionysius presents him as a master of the ideal mean (section 3.5). In On Demosthenes , Dionysius goes to great lengths to demonstrate Demosthenes’ superiority over the principal exponents in each of the three types of word selection: the critic considers Demosthenes better than Lysias in the simple style, better than Thucydides in the elevated style, and better than Plato and Isocrates in the mixed style. Dionysius submits, for instance, that Lysias’ simplicity is well-suited for expositions of facts, but that his style grows ‘faint’ ( ἀμυδρός ) and ‘weak’ ( ἀσθενής ), and eventually ‘burns out’ ( ἀποσβέννυται ) in other parts of his speeches; Demosthenes, conversely, possesses all of Lysias’ essential virtues, but he can also draw on a large reservoir of additional virtues.75 Next, Dionysius finds fault with Thucydides’ unrestrained style which often compromises the indispensable virtue of clarity; Demosthenes, however, at times uses the same kind of extraordinary, striking language as the historian, but without sacrificing the perspicuity of his words.76 In the mixed style, lastly, Dionysius argues that Plato and Isocrates aim to create a style that occupies a middle ground

73 See esp. section 3.2, tables 4 and 5 below. 74 Dion. Hal. Dem. 34.5, discussed in section 3.5 below. 75 Dion. Hal. Dem. 13.8. For the criticism of Lysias, see also Lys. 13.4, 19.5–6. The comparison between Lysias and Demosthenes is reminiscent of Cic. Opt. gen. 10: ‘Demosthenes could certainly speak calmly, but Lysias perhaps not with passion’ ( ita fit ut Demosthenes certe possit summisse dicere, elate Lysias fortasse non possit ). Longinus also seems to tap into a distinction between essential and additional virtues, cf. Long. Subl. 32.8 (= Caecilius T45 Woerther = IV 64 Ofenloch): Caecilius praised Lysias for being ‘faultless’ ( ἀναμάρτητος ) and ‘pure’ ( καθαρός ), but Longinus insists that a sublime author should aspire at greater, more risky, things. 76 Dion. Hal. Dem. 10.1–3: alluding to Thuc. 1.22.4, Dionysius playfully remarks that Demosthenes not only, like Thucydides, aimed ‘for permanent literary value’ ( εἰς ἀνάθημα καὶ κτῆμα ), but also ‘for practical use’ ( εἰς χρῆσιν ). On Thucydides’ unclarity, see also Dion. Hal. Thuc. 51.1–4 (on his intricate syntax), ibid. 55.1–2, Cic. Orat. 30 (on his rough arrangement of words), Long. Subl. 22.3 (on hyperbaton). Cf. De Jonge (2017), who reviews Dionysius’ assessment of Thucydides in the light of his program of rhetorical education and his Roman audience, which included the historian Q. Aelius Tubero.

CHAPTER TWO between Lysianic simplicity and Thucydidean elevation, but that they often fail to achieve a proper mean; Demosthenes, however, always hits the mark.77 To Dionysius’ mind, then, Demosthenes is the only prose author ever who holds a perfect equilibrium in the twisting kaleidoscope that is Attic prose. Well-proportioned versatility is not the only stylistic feature that Dionysius associates specifically with Demosthenes: the Greek scholar also argues that no orator has more talent than Demosthenes for the real-life oratory of the courts and the assemblies, that is, for forensic and deliberative speeches: the orator’s never uses ornaments merely to display his verbal artistry, but he always keeps his eye ‘on practical purposes’ ( ἐπὶ τὸ χρήσιμον ). 78 Dionysius pits the force of Demosthenes’ On the Crown , his favorite Demosthenic speech, against the ceremonial tranquility of the funeral oration in Plato’s Menexenus . Punning on Plato’s cave allegory, Dionysius compares On the Crown to ‘true visions’ ( ἀληθιναὶ ὄψεις ), while Menexenus is said not to surpass the epistemological level of mere ‘images’ (εἴδωλα ). Next, Demosthenes’ words are said to be like ‘weapons of war’ ( πολεμιστήρια ὅπλα ) and ‘bodies developed by hard work in the sun’ ( ἐν ἡλίῳ καὶ πόνοις τεθραμμένα σώματα ), whereas Plato’s words are more like ‘ceremonial weapons’ ( πομπευτήρια ὅπλα ) and ‘bodies that pursue a life of ease in the shade’ ( σκιὰς καὶ ῥᾳστώνας διώκοντα σώματα ).79 In addition, Dionysius argues that ‘one would not be far wrong to compare the style of Plato to a country spot full of flowers, which affords a congenial resting-place and passing delectation to the traveler, whereas that of Demosthenes is like a field of rich and fertile land, which yields freely both the necessities of life and the extra luxuries that men enjoy’. 80

77 On Plato’s mistakes in applying the mixed style, see esp. Dion. Hal. Dem. 5–7, repeated with some modifications at Pomp. 2.1; for Isocrates’ mistakes, see esp. Dion. Hal. Dem. 18.3–9, cf. Isoc. 13.4. See section 3.5 below on the importance of ‘appropriateness’ ( τὸ πρέπον ) in Demosthenes’ mixed style. 78 Dion. Hal. Dem. 32.2. Dionysius stresses Demosthenes’ focus on ‘usefulness’ ( χρῆσις , χρήσιμον ), for ‘reality’ (ἀληθινόν ) and for the actual ‘contests’ of the law courts and the assemblies ( ἀγῶνες , ἐναγώνιοι λόγοι ): see e.g. ibid. 10.3 (discussed in n. 76 above), 18.4, 20.3, 21.4, 32.3, 34.7, 45.1. Ooms and De Jonge (2013) 100–102 discuss the connections between Dionysius’ ἐναγώνιος λόγος and Aristotle’s λέξις ἀγωνιστική (Rh. 3.12): both terms refer to forensic and deliberative oratory, excluding epideictic speeches. Yunis (2019) 102–104 notes that Dionysius praises Demosthenes ‘for mixing the goals of immediate agonistic victory and enduring literary fame’. 79 Dion. Hal. Dem. 32.1: Dionysius quotes Pl. Menex. 237c–238a, 246c–248c, and Dem. De cor. 199–208. For the image of soldiers/athletes, cf. section 2.3.1 n. 44 above on Cleochares’ comparison between Demosthenes and Isocrates. For De cor. as Dionysius’ favorite speech by Demosthenes, see Comp. 25.25. 80 Dion. Hal. Dem. 32.2: Καί μοι δοκεῖ τις οὐκ ἂν ἁμαρτεῖν τὴν μὲν Πλάτωνος λέξιν εἰκάσας ἀνθηρῷ χωρίῳ καταγωγὰς ἡδείας ἔχοντι καὶ τέρψεις ἐφημέρους , τὴν δὲ Δημοσθένους διάλεκτον εὐκάρπῳ καὶ παμφόρῳ γῇ καὶ

HERO VERSUS ZERO

Dionysius’ point is that in the battles of real life Demosthenes is more useful than Plato: the former is better suited for the law-courts and the assemblies, while the latter is more at home in festivals and ceremonies. An even more dramatic picture of Demosthenes’ practical force emerges, when Dionysius compares him to Isocrates. 81 The Greek critic describes his feelings after reading Isocrates as follows: ‘I become serious and feel a great tranquility of mind, like those listening to libation-music played on reed-pipes or to Dorian or enharmonic melodies.’ When he picks up a speech by Demosthenes, however, he is anything but lulled to sleep: ‘I am transported, I am led hither and thither, (…) I feel exactly the same as those who take part in the Corybantic dances and the rites of Cybele the Mother-Goddess, and other similar ceremonies, whether it is because these celebrants are inspired by the scents, sights, or sounds or by the influence of the deities themselves, that they experience many and various sensations.’82 To be brief, Isocrates and Plato can be read at leisure, but Demosthenes compels his listeners and readers to feel the emotions that he expresses: he has the ability to overwhelm and even control his audience. Such ‘forcefulness’ ( δεινότης , ἰσχύς ) is extremely useful in swaying audiences and winning cases. 83 Finally, the third stylistic property that Dionysius considers typical for the style of Demosthenes is his knack for word arrangement, a topic that will be central in chapter four. Here, Dionysius deploys his threefold system of ‘types of word arrangement’ ( χαρακτῆρες τῆς συνθέσεως ), or three ‘harmonies’ ( ἁρμονίαι ), distinguishing between one type that is

οὔτε τῶν ἀναγκαίων εἰς βίον οὔτε τῶν περιττῶν εἰς τέρψιν σπανιζούσῃ. The combination of ‘necessities’ (ἀναγκαῖα) and ‘luxuries’ ( περιττά ) refers to the orator’s successful application of both the essential and the additional virtues of style; the word περιττός is often used by Dionysius in relation to the elevated style of word selection, cf. section 3.2 tables 2 and 5 below. 81 Note that in comparing Demosthenes to Plato, Dionysius compares a forensic speech (Dem. De cor. ) to an epideictic speech (Pl. Menex. ). In comparing Demosthenes to Isocrates, however, Dion. Hal. Dem. 22.1 submits that it does not matter what kind of speech by Isocrates one picks—forensic, deliberative or epideictic. 82 Dion. Hal. Dem. 22.1: Τὰ ἤθη σπουδαῖος γίνομαι καὶ πολὺ τὸ εὐσταθὲς ἔχω τῆς γνώμης , ὥσπερ οἱ τῶν σπονδείων αὐλημάτων ἢ τῶν Δωρίων τε καὶ ἐναρμονίων μελῶν ἀκροώμενοι . Ibid. 22.2–3: ἐνθουσιῶ τε καὶ δεῦρο κἀκεῖσε ἄγομαι , (...) διαφέρειν τε οὐδὲν ἑμαυτῷ δοκῶ τῶν τὰ μητρῷα καὶ τὰ κορυβαντικὰ καὶ ὅσα τούτοις παραπλήσιά ἐστι τελουμένων , εἴτε ὀσμαῖς ἐκεῖνοί γε εἴτε ἤχοις εἴτε τῷ δαιμόνων πνεύματι αὐτῷ κινούμενοι τὰς πολλὰς καὶ ποικίλας ἐκεῖνοι λαμβάνουσι φαντασίας . The Dorian and enharmonic melodies are associated with the sedate solemnity of the spondaic rhythm ( ― ―): see Aujac (1988) 168 n. 3. See Jonge (2012) 286–287 for the religious imagery and the ‘sublime’ language of this passage: ‘The emotional impact of Demosthenes’ speeches is here caught in what we might call ‘Longinian’ terminology.’ Cf. Long. Subl. 1.4 on being transported, ibid. 8.1, 8.4 and 15.1 on ecstasy ( ἐνθουσιασμός ), and ibid. 39.2 on Corybantic dancers. 83 On Dionysius’ forcefulness, see section 2.3.1 above.

CHAPTER TWO

‘smooth’ ( γλαφυρά ), another that is ‘rough’ ( αὐστηρά ) and a final one that is ‘well-mixed’ (εὔκρατος ): smooth composition serves to create ‘pleasure’ (ἡδονή ), rough composition can be used to produce ‘beauty’ ( τὸ καλόν ), while only the blended style succeeds in achieving both goals (section 3.2.2). 84 Again, Dionysius makes Demosthenes the all-round champion of the mixed style. In On the Arrangement of Words , accordingly, we read: ‘Demosthenes, in fact, is a sort of standard (ὅρος ) for both choice of words and beauty of arrangement.’ In On Demosthenes , likewise, Dionysius confidently states that ‘his arrangement of words is extraordinarily artistic and far superior to that of all other orators’. 85 To understand what it is about the orator’s arrangement that Dionysius admires, we should turn to the four main tools for effective composition that the critic recognizes: he argues that pleasure and beauty in arrangement can be achieved by means of ‘tone’ ( μέλος ), ‘rhythm’ ( ῥυθμός ), ‘variation’ (μεταβολή ) and ‘appropriateness’ ( τὸ πρέπον ).86 Needless to say, the latter two elements can be readily recognized in Demosthenes’ ability to adapt his style to suit any occasion: ‘In brief, he not only considered it necessary to vary the mixture of the styles of arrangement according to the individual requirements of his speeches and their different subject-matter, but he also saw that the constituent parts of the various forms of argument were of a different nature from one another, and tried to invest them with different styles, couching his aphoristic utterances in one kind of arrangement, his arguments in another and his examples in a different form again.’ 87 Moreover, variation and

84 For the two goals of word arrangement, see Dion. Hal. Comp. 10.2, 11.2, Dem. 47.2–4 with section 3.2, table 6 below. Donadi (1986) argues that Dionysius creates a novel approach to literary criticism by insisting on ‘pleasure’ ( ἡδονή ) and ‘beauty’ ( καλόν ), which applies to both visual and aural perception: for beauty and pleasure in sound, see esp. sections 4.3 and 4.4. Goudriaan (1989) 203–215 shows that the critic’s political and moral program can be closely linked to the two goals of word arrangement, esp. beauty: cf. section 4.5. 85 Dion. Hal. Comp. 18.15: Ὅρος γὰρ δή τίς ἐστιν ἐκλογῆς τε ὀνομάτων καὶ κάλλους συνθέσεως ὁ Δημοσθένης . Dem. 35.1: Περιττή τίς ἐστιν ἡ τῆς λέξεως τῆς Δημοσθένους ἁρμονία καὶ μακρῷ δή τινι διαλλάττουσα τὰς τῶν ἄλλων ῥητόρων . Note that Dionysius declares Demosthenes the champion of word arrangement in the category of oratory: at Comp. 18.13–14, he posits that Plato could rival Demosthenes on the subject of word arrangement. Cf. ibid. 19.12, where he praises Herodotus, Plato and Demosthenes in equal measure, submitting that the differences can be attributed to their different genres, i.e., history, dialogue, oratory. 86 See Dion. Hal. Comp. 11.1, Dem. 47.4. For μέλος as ‘tone’ instead of ‘melody’, see section 4.4 n. 64 below. 87 Dion. Hal. Dem. 46.1: Συνελόντι δ’ εἰπεῖν, οὐ μόνον παρὰ τὰς ἰδιότητας τῶν λόγων καὶ τὰς παραλλαγὰς τῶν ὑποθέσεων διαφόρους ᾤετο δεῖν ποιεῖσθαι τὰς κράσεις τῶν ἐν τῇ συνθέσει χαρακτήρων , ἀλλὰ καὶ παρ ’ αὐτὰ τὰ γένη τῶν ἐπιχειρημάτων τὰ συμπληρωτικὰ μέρη διαφόρους ἔχοντα τὰς φύσεις ὁρῶν διαλλαττούσαις κατασκευαῖς τῆς ἁρμονίας ἐπειρᾶτο κοσμεῖν, ἄλλως μὲν τὰς γνωμολογίας συντιθείς , ἄλλως δὲ τὰ ἐνθυμήματα , διαφόρως δὲ τὰ παραδείγματα .

HERO VERSUS ZERO appropriateness are presented as Demosthenes’ guiding principles in the areas of tone and rhythm. Under the heading of ‘tone’ ( μέλος ), Dionysius focuses mainly on the sound that results from the juxtaposition of final syllables and the subsequent initial syllables: clashes of vowels (e.g., εὐτυχοῦντα ὁρᾶν) or consonants (e.g., φοβερὸν προσπολεμῆσαι ) produce a rough sound, while combinations that can be pronounced without a pause (e.g., εἰ δέ τις ὑμῶν) are considered smooth (section 4.4). According to Dionysius, any excerpt from the orator’s speeches may serve to demonstrate his well-wrought variations of tone: the Greek critic underlines this point by claiming that he quotes samples from Demosthenes’ speeches that he selected ‘not deliberately, but at random’.88 As for rhythm, finally, Dionysius again ascribes to Demosthenes an impeccable sense of propriety: the orator is said to possess the ability to make his speeches resemble beautiful poems, but the rules of decorum prevent him from creating complete verses. This is an important point that the critic makes about the orator’s style: it is ‘rhythmical’ ( εὔρυθμος ) without being ‘in rhythm’ ( ἔρρυθμος ), it is ‘metrical’ ( εὔμετρος ) without being ‘in meter’ (ἔμμετρος ), holding a perfect middle ground between prose and poetry. 89 In addition,

88 Dion. Hal. Dem. 43.3: Οὐκ ἐξ ἐπιτηδεύσεως , ἀλλ ’ οἷς ἐνέτυχον . The collocations quoted here come from Dem. Olynth. II 22–23, which is discussed at Dion. Hal. Dem. 43.3–13 in order to verify ‘whether the structure is sometimes halting and broken up, sometimes coherent and compact, sometimes harshly grating on the ear, sometimes gently soothing, sometimes impelling hearers to emotion, sometimes leading gently on to moral seriousness, and producing many different effects in the actual arrangement’ ( εἰ τὰ μὲν ἀναβεβλημένας ἔχει τὰς ἁρμονίας καὶ διεστώσας , τὰ δὲ προσκολλώσας καὶ συμπεπυκνωμένας , καὶ τὰ μὲν ἀποτραχύνει τε καὶ πικραίνει τὴν ἀκοήν , τὰ δὲ πραΰνει καὶ λεαίνει , καὶ τὰ μὲν εἰς πάθος ἐκτρέπει τοὺς ἀκούοντας , τὰ δ’εἰς ἦθος ὑπάγεται , τὰ δ’ ἄλλας τινὰς ἐργάζεται καὶ πολλὰς διαφορὰς παρ ’ αὐτὴν τὴν σύνθεσιν ). Yet, Vaahtera (1997) demonstrates that Dionysius’ views about euphony and phonetics are often not pertinent to the passages that he cites to illustrate them: cf. section 4.4 n. 68 below. 89 For this point, see Dion. Hal. Comp. 11.25, 25.11–13, Dem. 50.7. Cf. Theon Prog. p. 71.7–11 Spengel, censuring Hegesias and other so-called Asian orators for applying a style that is both ‘in meter’ ( ἔμμετρος ) and ‘in rhythm’ ( ἔρρυθμος ): see n. 140 below. Dionysius attributes the criticism of overly metrical or rhythmical prose to Aristotle, cf. Arist. Rh. 3.8.3: ‘Therefore, prose must be rhythmical, but not metrical, otherwise it would be a poem; nor must this rhythm be rigorously carried out, but only up to a certain point’ (διὸ ῥυθμὸν δεῖ ἔχειν τὸν λόγον , μέτρον δὲ μή· ποίημα γὰρ ἔσται . ῥυθμὸν δὲ μὴ ἀκριβῶς· τοῦτο δὲ ἔσται , ἐὰν μέχρι του ᾖ). Transl. Kennedy (1991). In Comp. 25–26, Dionysius illustrates how prose and poetry may be similar: he quotes passages from Dem. Arist. and De cor. , arguing that the orator’s sentences constitute near-complete verses. De Jonge (2008) 294–295, however, rightly notes that ‘Aristotle would probably not have approved of Dionysius’ analysis of Demosthenes’ prose into almost complete verses’, suggesting that ‘Dionysius uses Aristotle as an authority for his own theories, albeit the philosopher’s views were actually rather different’. For the dichotomy between ‘rhythmical’ and ‘in rhythm’ in Latin, cf. Cic. De or. 3.184–185, Orat. 195, 220 with section 1.7 n. 151 above.

CHAPTER TWO

Dionysius insists that in Demosthenes’ speeches ‘the most frequent rhythms will be the manly, dignified and noble ones, rarely the loose rhythm of the Ionian choral dance’. In the former category, the critic typically lists those rhythmical feet that are dominated by long syllables, such as the spondee ( ― ―), the molossus ( ― ― ―), the bacchius ( ― ― ᴗ) and the hypobacchius ( ᴗ ― ―), while the latter category mostly features those feet that are dominated by short syllables, such as the pyrrhic ( ᴗ ᴗ), the tribrach (ᴗ ᴗ ᴗ) and the amphibrach ( ᴗ ― ᴗ). 90 Interestingly, modern scholarship confirms that Demosthenes eschewed successions of more than two short syllables: this rule, which is usually called ‘Blass’ law’ after its discoverer, seems to apply to no other fourth-century orator and it can hence be used as a negative criterion of authenticity for Demosthenes’ speeches.91 Thus, Dionysius praises Demosthenes’ ability to adapt his oratory to suit any occasion: in his view, the Athenian is a champion of the well-balanced, mixed style in the selection of his words and in their arrangement. Dionysius especially admires the orator’s sense of ‘appropriateness, which touches the stars in Demosthenes’ ( τὸ πρέπον ὃ τῶν ἄστρων ψαύει παρὰ Δημοσθένει ).92 This talent is particularly useful in the harsh reality of the law-courts and the assemblies, where any failure to observe the rules of decorum can have disastrous

90 Dion. Hal. Comp. 43.13: Καὶ τῶν ῥυθμῶν πολλαχῇ μὲν τοὺς ἀνδρώδεις καὶ ἁξιωματικοὺς καὶ εὐγενεῖς, σπανίως δέ που τοὺς ὑπορχηματικούς τε καὶ Ἰωνικοὺς καὶ διακλωμένους . For a complete overview of Dionysius’ noble and ignoble rhythms, see section 2.4.3 n. 146 on Comp. 17. Cf. also section 4.4 below on the stylistic applications of long syllables in the so-called ‘rough’ type of word arrangement. Note, however, that Dionysius considers dactyls ( ― ᴗ ᴗ) and the anapests ( ᴗ ᴗ ―) noble, while he describes the amphibrach ( ᴗ ― ᴗ), which has the same number of short syllables, as unpleasant and effeminate: cf. Gentili (1990) on the role of irrationality in Dionysius’ views on rhythm and meter. Confusingly, Dionysius (e.g. Comp. 17.7) uses the term ‘choree’, or ‘dance-like’ ( χορεῖος ) to refer to the tribrach ( ᴗ ᴗ ᴗ), while Cicero (e.g. Orat. 193) uses the same term ( choreus ) to refer to the trochee ( ― ᴗ). Apparently, various rhythmical feet, esp. those dominated by short syllables, could be associated with dancing: cf. the pyrrhic ( ᴗ ᴗ), so named after the ‘pyrrhic war-dance’ (πυρρίχη ). Arist. Rh. 3.8.4 posits that the trochee is ‘rather like the cordax’ ( κορδακικώτερος ), referring to the comedic dance ( κόρδαξ ), probably giving rise to Cicero’s use of the term choreus for the trochee. Incidentally, Cicero (e.g. Orat. 193) refers to the tribrach as ‘trochee’ ( trochaeus ). 91 Blass (1893) 105–112 noted that a speech is probably not composed by Demosthenes, if it has frequent tribrachs ( ᴗ ᴗ ᴗ). This ‘law’ has often been reaffirmed by later authors, but it should probably be considered a strong tendency rather than a rigid law: cf. e.g. Adams (1917), Skimina (1937) 106–136 and esp. McCabe (1981), who has verified the validity of Blass’ law on the basis of computer analyses. There are no extant ancient authors who explicitly mention this feature of Demosthenes’ style (although Dionysius hints at it), while Aelius Aristides (second century AD) is the only other known ancient writer who applies it: McCabe (1981) 20. 92 Dion. Hal. Dem. 34.5. The reference to stars touches on the rhetoric of the sublime: see section 3.5 n. 118.

HERO VERSUS ZERO consequences for the orator or his client. It is therefore, in Dionysius’ view, that Demosthenes does not use such extravagant gimmicks as dance-like rhythms, but he rather applies a cadence that reflects the seriousness of the case at hand.

2.3.4 Summary: the Principal Stylistic Virtues of Demosthenes To summarize, I have observed several similarities between Dionysius’ and Cicero’s discussions of Demosthenes’ style: both authors claim that he is the best exponent of Greek prose, while they also submit similar reasons for doing so. From the material that I have examined above, three principal stylistic virtues emerge that both the Roman rhetorician and the Greek critic adduce to substantiate their claim that Demosthenes should be awarded the crown of eloquence. In a word, the common Greek and Roman discourse on the style of Demosthenes in Late-Republican and Augustan Rome can be summarized in the following three points:

1. Cicero and Dionysius stress Demosthenes’ stylistic versatility. According to Cicero, the orator’s repertoire spans all three stylistic registers, allowing him to speak appropriately on every subject. Dionysius argues that Demosthenes is a champion of the mixed style, which is an appropriate combination of the best elements from the simple and elevated styles. The orator’s oeuvre was thought to offer a comprehensive overview of stylistic theory put into practice, which makes him the ideal object of imitation for Greek and Roman orators alike.

2. Demosthenes was admired mainly as a practical orator, whose style is particularly effective in forensic and deliberative speeches. Cicero notes that his forcefulness is extremely useful on the oratorical battleground of the Forum. Dionysius compares the usefulness and the astonishing force of Demosthenes’ style, likewise, to weapons of war, fertile fields, highly trained bodies, and the ecstatic rites of Cybele. It was thought, then, that Greek and Roman orators could overwhelm and control their audience by summoning the power of Demosthenic thunderbolts.

3. According to Cicero and Dionysius, Demosthenes exhibited an unrivalled talent for word arrangement. Cicero, who does not go into much detail about the orator’s style of arrangement, attributes the force of his oratory to a large extent to his attention to rhythm. Dionysius, who offers an extensive discussion of the orator’s verbal

CHAPTER TWO

arrangements, also praises his use of rhythm, which allowed him to compose speeches that are not only admirable as artistic masterpieces resembling beautiful poetry, but also to be reckoned with in serious oratory.

In the next chapters, I will examine the differences between the stylistic discussions in Cicero’s rhetorical works and Dionysius’ critical essays that I already briefly signaled above. I will, for instance, zoom in on the different motivations that underlie their three-style systems (sections 3.4 and 3.5), and I will pay attention to their diverse evaluations of musical arrangement (section 4.5). In addition, we will also see that Cicero and Dionysius have different ideas about the practical application of their doctrines: it will appear that stylistic theory is closely connected with the politics of Late-Republican and Augustan Rome (sections 5.5 and 5.6). The next section, however, will dwell a little longer on the common critical framework of our authors by turning to their discussions of Hegesias of Magnesia, whom they both consider to be Demosthenes’ stylistic antipode.

2.4 How Not To Compose Great Prose: Hegesias as the Prototype of Degeneracy According to Dionysius, Hegesias is a ‘high-priest of humbug’ ( τῶν λήρων ἱερεύς ). 93 This kind of abuse is typical for the extreme scorn that the Hellenistic orator and historian receives from virtually every extant ancient source that discusses his style. 94 Although the absolute peak of anti-Hegesias sentiment, exhibited by Cicero, Dionysius, Strabo, Aelius Theon and Longinus, lies in the first centuries BC and AD, he was already severely censured by the historian and geographer Agatharchides of Cnidus (fl. mid-second century BC). 95 In this

93 Dion. Hal. Comp. 4.11. 94 The only clear-cut exception is the Roman scholar Varro (116–27 BC): cf. Cic. Att. 12.6.1 cited in section 2.1 n. 6 above. Rutilius Lupus (late first century BC) and Philodemus are not obviously hostile to Hegesias. Rutilius translates various Greek passages into Latin to illustrate the figures of speech catalogued in his excerpt of On Figures of Speech by the Greek rhetor Gorgias (first century BC): two books survive, Quint. Inst. orat. 9.2.102 knew four books. Rutilius has examples from Hegesias of epanaphora ( ἐπιβολή , Rut. Fig. 1.7 Halm = Hegesias F27 Jacoby), varied construction ( ἀλλοίωσις , Rut. Fig. 2.2 Halm = Hegesias F28 Jacoby) and feigned doubt (ἀπορία , Rut. Fig. 2.10 Halm = Hegesias F29 Jacoby): Rutilius refers to Hegesias in the same way as to Attic orators like Demosthenes. For Philodemus, see Rhet. 4 col. 21.15–25 (= p. 180 Sudhaus) in section 2.2 n. 25. 95 Str. Geogr. 14.1.41, Theon Prog. p. 71.7–11 Spengel, Long. Subl. 3.2: see section 2.2 n. 14, 24 and 25 above. The available evidence for Agatharchides’ life is meagre, but from a few testimonies (Str. Geogr. 14.2.15 = Agatharch. T1 Jacoby; Phot. Bibl. cod. 213, 171a6–b17 = Agatharch. T2 Jacoby) and a couple of autobiographical references in his work On the Red Sea (quoted at Phot. Bibl. cod. 250, 460b3–13 = Agatharch.

HERO VERSUS ZERO section, I will first examine the discussion of Hegesias’ style by the latter author (section 2.4.1). Next, we will see that this earlier criticism foreshadows the harsh judgments in both Cicero’s rhetorical works (section 2.4.2) and Dionysius’ critical treatises (section 2.4.3): as their designated antimodel of style, Hegesias seems to fail on precisely the same points on which Demosthenes is said to have triumphed. Thus, we will see that Hegesias falls outside of the scope of the threefold stylistic classifications of our authors, that his style is considered utterly unpractical, not to be used in real-life oratory, and that he is thought to possess no talent whatsoever for appropriate word arrangement. In brief, Hegesias is set up to be Demosthenes’ polar opposite.

2.4.1 Agatharchides of Cnidus on Hegesias’ Style The only discussion of Hegesias’ style before the first century BC comes from an excerpt from the last book of Agatharchides’ five-volume geographical, ethnographical and historiographical work On the Red Sea .96 The author’s interest in Hegesias is sparked, when he ponders the proper approach to describing the calamities of others: ‘Many public orators and poets have been at a loss as to how a person whose situation is free from danger ought properly (πρεπόντως ) to recount the extreme misfortunes that have befallen some men.’ 97 As

T3 Jacoby), we can gather that he was probably born before 200 BC, he stayed at the court of Ptolemy VI Philometor in Alexandria, he was exiled around 145 BC by Ptolemy VIII, and he spent the remainder of his life in Athens: see Burstein (1989) 12–18, id. (2012). On Agatharchides’ works, esp. Erythr. , see section 2.4.1 n. 96. 96 Phot. Bibl. cod. 250 preserves two large excerpts from On the Red Sea (Περὶ τῆς Ἐρυθρᾶς θαλάσσης ), one from the first book and one from the fifth book. The relevant passage on Hegesias is Agatharch. Erythr. 5.21 (= Agatharch. p. 119.14–122.24 Müller, Geographi Graeci Minores 1 = Phot. Bibl. 445b36–447b5), containing Hegesias T3, F6–19. Erythr. is Agatharchides’ third major work, written after his large treatises On Affairs in Europe (Τὰ κατὰ τὴν Εὐρώπην ) in ten books and On Affairs in Asia (Τὰ κατὰ τὴν Ἀσίαν ) in forty-nine books. He probably worked on it during his stay at the Ptolemaic court until ca. 145 BC: Agatharchides himself claims (at Phot. Bibl. 460b3–13) that he left Erythr. unfinished on account of his old age, and because he lost access to the Ptolemaic archives after his exile (cf. previous note). The accounts on the Red Sea in Diod. Sic. 3.12–48 and Str. Geogr. 16.4.5–20 are derived from Agatharchides. There is some discussion as to whether Erythr. is a separate work or rather an integral part of On Affairs in Asia : see Marcotte (2001) for the latter interpretation, and Engels (2004), who gives an overview of modern scholarship on the work. Burstein (2012) offers a commentary to Jacoby’s collection of fragments. 97 Agatharch. Erythr. p. 119.14–17 Müller (= Phot. Bibl. 445b38–41): Ὅτι πολλοὶ καὶ τῶν πολιτικῶν ἀνδρῶν καὶ τῶν ποίημα γεγραφότων διηπορήκασι πῶς τὰς ὑπερβαλλούσας ἐνίοις ἀκληρίας τὸν ἐκτὸς τῶν κινδύνων κείμενον πρεπόντως ἐξαγγελτέον . Cf. Dion. Hal. Dem. 22.4 on Demosthenes’ ability to make later generations feel the emotions that he meant to pass onto his contemporaries: see section 2.2 n. 21 above.

CHAPTER TWO a case in point, Agatharchides turns to the destruction of Olynthus (348 BC) by the Macedonian king Philip II, and the sack of Thebes (335 BC) by his son Alexander: he quotes twelve (or perhaps fourteen) small samples from Hegesias’ works as typical examples of how not to describe such tragedies.98 According to the author, the Magnesian misguidedly strives for verbal artistry, where pathos is required: ‘Hegesias, who often commemorated the destruction of these cities, is paltry ( εὐτελής ). For, not wishing to speak in a manner appropriate to the occasion (τοῖς καιροῖς οἰκείως ) but compulsively seeking to display his preciosity (κομψότης ) in dealing with a harsh topic, he achieves, to some extent, his personal goal, but without regard for the dignity of his subject.’ 99 Agatharchides, then, posits that Hegesias’ style is unfit for serious matters: this point will return in the works of Cicero and Dionysius, who, as we have already seen, stress the importance of appropriateness in their analysis of Demosthenes. In addition, there is another aspect in the passage from On the Red Sea that seems to anticipate the stylistic discourse of Late-Republican and Augustan Rome: Agatharchides not only censures Hegesias, but also ‘the followers of Hegesias’ ( οἱ περὶ Ἡγησίαν ) and a certain Hermesianax, while he praises the Attic orators Demosthenes, Aeschines and Stratocles. 100 According to Agatharchides, the

98 There is no consensus about the origins of the citations: for a discussion, see Prandi (2016). As Susemihl (1892) 464 and Jacoby (1930) 529 assumed, the fragments may derive from a historiographical work on Alexander the Great: they can easily be imagined to belong to a larger narrative, and Dion. Hal. Comp. 18.22 quotes a thematically related passage about Alexander’s siege of Gaza (332 BC) ‘from his history’ ( ἐκ τῆς ἰστορίας ). Yet, Sternbach (1930) 43 may be right in ascribing the fragments to various rhetorical works: Hegesias often uses direct speech, Agatharchides refers to his ‘speeches’ ( λόγοι ), and he calls him a ‘sophist’ (σοφιστής ). Staab (2004) 136–146 suggests that the fragments come from his satirical dialogue The Lovers of Athens : cf. section 2.2 n. 2 above. Phot. Bibl. 446b35–39 (= Hegesias F25–26 Jacoby) quotes two sentences of ambiguous authorship: Jacoby (1930) 531 and Staab (2004) 146 are rightly hesitant to attribute them to either Hermesianax (who is quoted immediately before, cf. n. 100 below) or Hegesias (whose style is Agatharchides’ main focus throughout the passage). It is equally possible that they come from another, unnamed Asian author, as Sternbach (1930) 44 suggests. 99 Agatharch. Erythr. p. 119.33–120.3 Müller (= Phot. Bibl. 446a16–21): Ἡγησίας μὲν οὖν πολλάκις τῆς ἀπωλείας μεμνημένος τῶν πόλεων εὐτελής ἐστιν . Ὁ γὰρ μὴ θέλων τοῖς καιροῖς οἰκείως διαλέγεσθαι , ζητῶν δὲ ἐξ ἀνάγκης ἐν αὐστηρῷ πράγματι κομψότητα διαφαίνειν , τοῦ μὲν ἰδίου ζηλώματος ἐπὶ ποσὸν τυγχάνει , τῆς δὲ τῶν ὑποκειμένων ἀξίας οὐ στοχάζεται . For Hegesias’ κομψότης , cf. Dion. Hal. Comp. 4.11 ( μικρόκομψον ). 100 See Agatharch. Erythr. p. 121.12–15 Müller (= Phot. Bibl. 446b33–36) for Hermesianax, see ibid. 121.38– 122.22 (= Phot. Bibl. 447a17–447b3) for Demosthenes, Aeschines and Stratocles. Stratocles (fl. ca. 300 BC) was usually criticized as a politician who contributed to the downfall of the Athenian democracy after Demosthenes, esp. due to his emphatic support of Demetrius Poliorcetes: cf. Ath. 13.596, Plut . Demetr . 12.1, Diod. Sic.

HERO VERSUS ZERO former have described the fall of Olynthus and Thebes ‘in an allegorical manner (ἀλληγορικῶς) and through extravagant expressions (περιττῶς)’, whereas the latter have spoken ‘in a more dignified way, maintaining the normal and proper meanings of the words, even while dealing with terrible matters’.101 Although Agatharchides himself does not use the labels ‘Attic’ and ‘Asian’, his analysis is reminiscent of the later discussions about Attic and Asian style in Rome, which typically connect Atticism with purity, and Asianism with abundance (section 5.3). Still, we should not consider the Agatharchides a pioneer of Atticism: admiration for the Attic dialect and Athenian culture was common throughout the Hellenistic era, and, as we will see, there did not exist a unified centuries-spanning school of Atticism with actual, identifiable adherents (section 5.2). 102 Why does Agatharchides consider Hegesias’ words ill-suited to describe the tragic fates of Olynthus and Thebes? As we have seen, the geographer specifically takes offense at the ‘preciosity’ ( κομψότης ), that is produced by Hegesias’ overabundant use of fancy, figurative language. Because of this vicious tendency, he is, according to the geographer, unable the express the emotion ( πάθος ) that is appropriate for the description of human suffering: Agatharchides divides his quotes from Hegesias into three groups that each illustrate one way in which the Magnesian destroys the genuine emotion that his subject requires. First of all, his extravagant expressions are often hard to understand, as can be seen in the examples below: 103

− F6 Jacoby: Ὄνομα κατελάβομεν πόλιν καταλιπόντες . (‘We have earned a name by abandoning a city.’)

20.46.2, and, for a nuanced modern discussion, Bayliss (2011) 152–186. The identity of Hermesianax, to whom Agatharchides attributes a Praise of Athens , is uncertain. He is often identified with the Alexandrian poet from Colophon (fl. 300–250 BC), who composed the three-book elegiac poem Leontion , but his appearance ‘in the midst of a long rant about the orator-historian Hegesias is slightly odd’: see Tuplin (2010). Another candidate is a certain Hermesianax of Cyprus, who wrote a Phrygian Affairs (Φρυγιακά ), mentioned at [Plut.] Fluv. 12.4. It is equally possible, however, that the references is to another, hitherto unknown Hermesianax. 101 Agatharch. Erythr. p. 119.25–29 Müller (= Phot. Bibl. 446a8–12): Εἰρήκασιν οὖν περὶ τούτου τοῦ πράγματος ἄλλοι μὲν ἀλληγορικῶς τῷ τρόπῳ καὶ ταῖς διαλέκτοις , ὡς δοκοῦσι , περιττῶς, οἱ δ’ ἐμβριθέστερον , τὰ συνήθη καὶ τὰς κυριολογίας ἐν τοῖς δεινοῖς οὐ πεφευγότες . 102 On account of Agatharchides’ praise of Attic orators and his rejection of Hegesias, Norden (1898) 127 thinks that Atticism originated ca. 200 BC; see, however, the criticism of this view in Radermacher (1899a) 351–360, Wilamowitz (1900) 28, and Wisse (1995) 74–75 and section 5.2 n. 21 below. 103 Agatharch. Erythr. p. 120.3–21 Müller (= Phot. Bibl. 446a22–32).

CHAPTER TWO

− F7 Jacoby: T ὸν γὰρ μέγιστα φωνήσαντα τόπον ἄφωνον ἡ συμφορὰ πεποίηκε . (‘For the disaster has made the place that has rung the loudest speechless.’) − F8 Jacoby: Ἐκ μυριάνδρου πόλεως ἐξῆλθον , ἐπιστραφεὶς δ’ οὐκέτ ’ εἶδον . (‘I left a city of ten thousand men, but when I turned around, I no longer saw it.’)

These sentences contain witty puns that are based on assonance ( κατελάβομεν , καταλιπόντες ), cognates ( φωνήσαντα , ἄφωνον ) and paradox (a populous city vanishing into thin air) respectively: Hegesias’ words are all the more puzzling for us, as we have little information about their original context.104 According to Agatharchides, such riddles have no place in the treatment of woes: ‘This makes no emotional impact at all but concentrates attention on his meaning and causes a person to struggle to understand what it is that he is talking about. For, wherever one creates uncertainty about the meaning, there one weakens a discourse.’ Rather, ‘a person uttering a lament must eschew witticisms (ἀστεϊσμοί ) and clearly indicate the event with which the emotion is connected, if he intends not merely to adorn his speech with fine phrases (διακοσμεῖν τῇ λέξει ) but to grasp the cause of the misery’. 105 Secondly, Agatharchides not only posits that Hegesias fails to make his meaning clear, but also that he is unable to understand ‘how to make the emotion (πάθος ) visible through vivid description’ ( ἐνάργεια ).106 Indeed, the idea that historical narrative should be vivid and full of emotion appears to have been fundamental to Agatharchides’ conception of his own work as a historian: he stresses the didactic force of ‘vivid description which teaches the audience about the matter at hand’ ( ἡ διδάσκουσα τὸ πρᾶγμα ἐνάργεια ) through emotional

104 Staab (2004) 139–143 offers a discussion of various ways to understand these sentences. According to him, F6 possibly refers to the fact that only the name of the city (probably Thebes) exists after its destruction: ‘Nur noch einen Namen bekamen wir zu fassen, nachdem wir die Stadt verlassen haben.’ Alternatively, Hegesias may have cleverly combined the phrases πόλιν καταλαβεῖν (‘to conquer a city’) and ὄνομα λαβεῖν (‘to take on a name’) as a cynical utterance from a vanquished speaker: ‘Einen Namen haben wir bekommen, indem wir eine Stadt zurückgelassen haben.’ Staab submits that the ‘loud ringing’ in T7 refers to the vociferous resistance of Thebes against Alexander. F8, lastly, addresses the destruction of Olynthus, as Agatharchides himself indicates. 105 Agatharch. Erythr. p. 120.5–9 Müller (= Phot. Bibl. 446a23–26): Τοῦτο πάθος μὲν οὐδαμῶς ποιεῖ, ἀλλὰ συνάγει πρὸς τὴν ἔμφασιν καὶ ποιεῖ ζητεῖν τί λέγει . Οὗ γὰρ ἄν τις τῇ γνώμῃ δισταγμὸν ἐνεργάσηται , ἐκεῖσε ἀπὸ τοῦ λόγου τὸ δεινὸν ἀφῆκε . Ibid. 120.18–21 Müller (= Phot. Bibl. 446a34–37): Δεῖ δὲ τὸν οἰκτιζόμενον ἀφέντα τοὺς ἀστεϊσμούς , τὰ πρᾶγμα σημαίνειν ᾧ οἰκείωται τὸ πάθος , εἰ μέλλοι μὴ τῇ λέξει διακοσμεῖν, ἀλλὰ τῷ τῆς συμφορᾶς αἰτίῳ προσεδρεύειν . 106 Agatharch. Erythr. p. 120.45–46 Müller (= Phot. Bibl. 446b18–19): Πῶς τὸ πάθος ὑπὸ τὴν ὄψιν ἀγάγοι διὰ τῆς ἐναργείας .

HERO VERSUS ZERO engagement rather than intellectual reflection. 107 According to Agatharchides, Hegesias utterly fails in this respect: his extravagant language does not contribute to bringing the gruesome destruction of Thebes and Olynthus before the eyes of his readers and listeners. Compare the following quotes that Agatharchides adduces: 108

− F9 Jacoby: Ἀλέξανδρε , καὶ τὸν Ἐπαμινώνδαν νόμισον , ὁρῶντα τὰ λείψανα τῆς πόλεως , παρόντα μοι συνικετεύειν . (‘Alexander, imagine also Epaminondas as present, and, after seeing the remains of the city, joining with me in supplication.’) − F10 Jacoby: Βασιλικῇ μανίᾳ προσπταίσασα πόλις τραγῳδίας ἐλεεινοτέρα γέγονε . (‘Struck by royal madness, the city has become more piteous than a tragedy.’) − F11 Jacoby: Τί δεῖ λέγειν Ὀλυνθίους καὶ Θηβαίους , οἷα κατὰ πόλεις ἀποθανόντες πεπόνθασιν; (‘Why should one mention the Olynthians and the Thebans, how they suffered, dying in their cities?’) − F13 Jacoby: Αἱ δὲ πόλεις αἱ πλησίον ἔκλαιον τὴν πόλιν , ὁρῶσαι τὴν πρότερον οὖσαν οὐκέτ ’ οὖσαν . (‘The nearby cities bewailed the city, seeing that what formerly existed no longer exists.’)

Agatharchides claims that Hegesias is unable to address the horror of his topic: taking the rhetorical question (in F11) au serieux, the geographer for instance complains that Hegesias is not even sure if it is at all necessary to commemorate the fate of the Olynthians and the Thebans. As for the other fragments, Agatharchides complains that Hegesias goes out of his way to create extraordinary phrases without any concern for the vehement emotion that they ought to convey. In the appeal to Alexander, for instance, ‘the request is puerile (μειρακιῶδες ) and the metaphor is harsh, but the grimness of the deed remains unaddressed’. In comparing

107 Agatharch. Erythr. p. 122.12–13 Müller (= Phot. Bibl. 447a35–36). See Zangara (2008) for a discussion of ‘vivid description’ ( ἐνάργεια ) as one of several ways in which Hellenistic historians make the events that they describe visible. See also Allan, De Jong and De Jonge (2017) on ἐνάργεια (‘the power of bringing the things that are said before the senses of the audience’) in Greek literary criticism and its links with the modern notion of immersion. As for Agatharchides, Gallo (2011) argues that he aimed to arouse emotions in order to get his moral messages across, while Maier (2018) posits that he uses vivid description as a means to make his audience experience history as an eyewitness, making ἐνάργεια a substitute for the author’s autopsy. 108 Agatharch. Erythr. p. 120.21–121.3 Müller (= Phot. Bibl. 446a37–b25). For (the rhythm of) F12, the longest citation of Hegesias in Agatharchides, see section 2.4.2 below. For a thoughtful reconstruction of the original contexts of these passages, see Staab (2004) 143–145. Epaminondas, referred to in F9, is a Theban leader (first half of the fourth century BC), who once entertained Philip II: cf. Diod. Sic. 15.39.2, 17.2.2 and Cic. Tusc. 1.4.

CHAPTER TWO the destruction of cities with a fictitious drama, likewise, ‘he appears to have created anything except what would be appropriate for a skilled speaker (τὸ καθῆκον σοφιστῇ), since it does not touch the subject at all.’ A similar point can be made about the last fragment, in which Hegesias uses a crude euphemism to refer to the ruin of the Greek cities: ‘What formerly existed, no longer exists’. Hence, Agatharchides concludes that Hegesias is unable to arouse pity: ‘If, therefore, someone spoke these sentences to the Thebans and Olynthians by way of consolation at the time of their conquest, they would, I think, have laughed at the author and, in a way, considered him more wretched than themselves.’ 109 This brings us to Agatharchides’ third series of quotes from Hegesias, which serve to show the ‘ultimate frigidity’ ( ψυχρότης ἐσχάτη ) that the Magnesian orator and historian commits. Frigidity arises, when an author, who aims at grandeur, overshoots the expression that is appropriate for the thought and ends up with an exaggerated, affected style. 110 Hegesias incurs the charge of frigidity on account of his intricate wordplays, as the following four citations aim to illustrate: 111

− F14 Jacoby: Δεινὸν τὴν χώραν ἄσπορον εἶναι τὴν τοὺς Σπαρτοὺς τεκοῦσαν . (‘Terrible is the fact that the land that bore the ‘Sown-men’ is unsown.’) − F15 Jacoby: Θηβαῖοι ἐν τῇ μάχῃ τῇ πρὸς Μακεδόνας ὑπὲρ τοὺς μυρίους ἀνετράπησαν . (‘The Thebans, more than ten thousand of them, were turned upside down in the battle against the Macedonians.’)

109 Agatharch. Erythr. p. 120.24–26 Müller (= Phot. Bibl. 446a40–b1): Αἴτημα μὲν μειρακιῶδες καὶ μεταφορὰ σκληρά , τὸ δὲ σκυθρωπὸν τῆς πράξεως ἄρρητον . Ibid. 120.28–29 Müller (= Phot. Bibl. 446b3–5): Πρὸς πᾶν μᾶλλον ἢ τὸ καθῆκον σοφιστῇ ἡτοιμάσθαι φαίνεται , ὅθεν οὐ λίαν ἅπτεται τοῦ προκειμένου . Ibid. 120.48–121.3 Müller (= Phot. Bibl. 446b21–25): Εἴ τις οὖν τοῖς Θηβαίοις καὶ τοῖς Ὀλυνθίοις παρ ’ αὐτὴν τὴν ἅλωσιν τὰς περιόδους ταύτας συναχθόμενος ἔλεγε , δοκοῦσιν ἄν μοι γελάσαι τὸν γράψαντα καὶ τρόπον τινὰ αὑτῶν ὑπολαβεῖν ἀθλιώτερον . For Hegesias’ puerility, cf. Cic. Brut. 287: puerile . For the ridiculousness of Hegesias’ style, cf. Dion. Hal. Comp. 18.28: οὐδὲ μετὰ σπουδῆς, ἀλλ ’ ἐπὶ χλευασμῷ καὶ καταγέλωτι . For Hegesias as a sophist, cf. Dion. Hal. Comp. 18.22 with section 2.1 n. 5 above. 110 Agatharch. Erythr. p. 121.12 Müller (= Phot. Bibl. 446b33). For frigidity as a vicious form of grandeur, see e.g. Demetr. Eloc. 114–127 (quoting Theophrastus); Dion. Hal. Isoc. 3.1, Dem. 20.6, 21.3, Thuc. 46.2, 48.3; Long. Subl. 3.4. Dion. Hal. Ant. orat. 2.4 connects frigidity to Asianism. For Hegesias’ frigidity, cf. Plut. Alex. 3.6. Van Hook (1917) notes that the modern term ‘frigidity’ (i.e., flatness, dullness, insipidity) does not correspond with the meaning of the Greek term ψυχρότης , which refers to extravagance; instead, he proposes ‘fustian’ as a translation; ‘frigidity’ seems to go better with the Latin frigidum . Cf. Gutzwiller (1969). 111 Agatharch. Erythr. p. 121.3–38 Müller (= Phot. Bibl. 446b25–447a16). See Staab (2004) 146–148 for a discussion of the possible meaning of these passages.

HERO VERSUS ZERO

− F16 Jacoby: Τῆς μὲν πόλεως κατασκαφείσης οἱ μὲν ἄνδρες παίδων συμφορὰς ὑπομένουσιν , αἱ δὲ γυναῖκες μετήχθησαν εἰς Μακεδονίαν , τὴν πόλιν θάψασαί τινα τρόπον . (‘After the city had been razed, the men put up with the miseries of their sons; the women were transported to Macedonia, having in a way buried the city.’) − F17 Jacoby: Ἡ δὲ φάλαγξ τῶν Μακεδόνων εἰσβιασαμένη τοῖς ὅπλοις ἐντὸς τείχους τὴν πόλιν ἀπέκτεινεν . (‘The Macedonian phalanx, having used its weapons to force its way within the walls, killed the city.’)

The first fragment above refers to the myth which relates how Cadmus planted a handful of dragon’s teeth in the barren Theban land, from which subsequently emerged a group of fully equipped warriors, known as the ‘Sown-men’ ( Σπαρτοί ). Agatharchides is particularly unhappy about the antithesis, which is derived from the pun on two cognates ( Σπαρτοί , ἄσπορος ), both connected to the verb ‘to sow’ ( σπείρω ): far from inciting pity or grief, such tongue-in-cheek plays on words will only arouse laughter. 112 The remaining examples are selected for their unfortunate figurative expressions: Hegesias, for instance, uses the verb ‘to turn upside down’ (ἀνατρέπω ), which evokes the image of capsizing ships, to describe the killing of ten thousand Thebans, a number which is probably exaggerated. ‘What a marvelous image,’ Agatharchides remarks cynically, ‘so many men unexpectedly turned upside down!’ 113 Similarly, Hegesias remarks that the women of Thebes ‘buried’ their city (F16), after it had been ‘killed’ by Macedonian invaders (F17). Unsurprisingly, the metaphors do not please the geographer: ‘There the grave of the city and here its death: all that is required is to add a funeral and compose an epitaph and the business would be complete!’ These ‘frigid’ quotes lead Agatharchides to diagnose their author with ‘madness’ ( μανία ).114

112 According to Agatharch. Erythr. p. 121.6–12 Müller (= Phot. Bibl. 446b27–33), Hegesias adapted this sentence from a phrase by Demosthenes of unknown origin, discussing the barrenness of Attica, which had been the birthplace of the cultivation of crops: Agatharchides submits that the Demosthenes’ antithesis is derived from the thought, while Hegesias derives it from the words. Agatharch. Erythr. 121.12–19 Müller (= Phot. Bibl. 446b33–39) cites a similar pun from Hermesianax, and two from an ambiguous source (cf. n. 98 above). 113 Agatharch. Erythr. p. 121.26–27 Müller (= Phot. Bibl. 447a5–7): Ὦ καλῆς ἐμφάσεως· ἄνθρωποι τοσοῦτοι παραλόγως ἀνατετραμμένοι . The word ἀνατρέπω can refer to anything that is turned upside down, such as men (e.g. Hom. Il. 6.164), ships (e.g. Pl. Lg. 906e), tables (e.g. Dem. Fals. leg. 198) and stomachs (e.g. Gal. 12.911). Diod. Sic. 17.14.1 and Ael. VH 13.7 report a grand total of 6.000 victims for the sack of Thebes in 335 BC. 114 Agatharch. Erythr. p. 121.34–36 Müller (= Phot. Bibl. 447a12–14): Ἐκεῖ μὲν ταφὴ πόλεως , ἐνταῦθα δὲ θάνατος· λοιπὸν ἐκφορὰν δεῖ προσθέντας ἐπιγράμματι χρήσασθαι , καὶ παντελὴς ἡ πρᾶξις . For the madness of Hegesias, cf. Dion. Hal. Comp. 18.22: διαφθορὰ τῶν φρενῶν.

CHAPTER TWO

To summarize, Agatharchides criticizes Hegesias’ obsession with fancy, extraordinary language, which renders his style obscure, devoid of vividness and frigid. Accordingly, he is, in Agatharchides’ view, utterly incapable of arousing the emotions that are appropriate to serious topics. As we have seen, the author of On the Red Sea reviews Hegesias’ style from the vantage point of his own historiographical enterprise: still, several elements of his harsh judgment will return in the rhetorical works of Cicero and in the critical essays of Dionysius. For all three detractors of Hegesias, inappropriateness seems to be a key word to understand his degeneracy.

2.4.2 Cicero on Hegesias’ Style Cicero focuses mainly on one aspect of Hegesias’ prose, that is, his rhythmical word arrangement. To be brief, the Roman rhetorician makes three interrelated points about the Asian author’s prose rhythm: (1) it is fragmentary, consisting almost exclusively of short phrases that create a staccato effect; (2) it has a strong poetical flavor, resembling an incessant stream of little verses; and (3) it is monotonous, as it applies the former two features throughout the entire discourse. By analyzing the rhythm in the three longest extant fragments of Hegesias’ works, we will see that Cicero’s assessment, despite its obvious bias, is founded in a serious stylistic analysis. Toward the end of his Orator , Cicero discusses the rhythmical errors that are typical for Asian orators. 115 He connects one of these mistakes specifically to the style of Hegesias, who in his view is always ‘in pursuit of short rhythms’ ( minutos numeros sequens ): ‘There are also those, who in the vicious manner which stems from Hegesias, by cutting and breaking up their rhythms ( infringendis concidendisque numeris ) fall into an insipid type of style (genus abiectum ) that resembles verselets ( versiculi ).’ According to Cicero, the Magnesian consistently pursues such chopped up sentences, refusing to vary their length: ‘There is no style better or stronger than to strike with phrases of two or three words, sometimes with single words, and at other times with several, in the midst of which comes sparingly the

115 Cic. Orat. 230–231. In addition to the error discussed below, Cicero mentions (1) orators who insert ‘certain meaningless words just to fill out the rhythm’ ( inania quaedam verba quasi complementa numerorum ), and (2) orators, like the brothers Hierocles and Menecles of Alabanda, whose sentences ‘all end in the same way’ ( omnia fere concludebantur uno modo ). The use of silly patch-words was associated with overly smooth word arrangements: cf. Sluiter (1997) 238–244 with e.g. Cic. Orat. 230, Demetr. Eloc. 55–58, and Dion. Hal. Comp. 22.5. Note that Cicero himself was associated with the vice of ending all of his sentences in a similar fashion, viz., by applying the cadence ― ᴗ ᴗ ᴗ | ― ― (e.g. esse videatur ): see section 4.2 n. 36 below.

HERO VERSUS ZERO rhythmical period (numerosa comprehensio ) with varying cadences (variis clausulis ). Hegesias perversely avoids this, and while he, too, tries to imitate Lysias, who is almost the equal of Demosthenes, he hops about ( saltat ), cutting his style into little fragments (particulae ).’ 116 Cicero’s casual reference to Demosthenes serves to highlight an important contrast between the Attic orator and the Asian author: whereas the former is the incontestable icon of stylistic versatility, Hegesias’ prose is utterly unvarying. 117 To make sense of what Cicero means, when he refers to ‘fragments’ ( particulae ) and ‘verselets’ ( versiculi ), we may turn to the only three surviving passages from Hegesias that contain more than one single sentence. Encouraged by Cicero’s remarks about Hegesias’ rhythmical prose, I have chopped the sentences in these texts up into short phrases and I have scanned the endings, or clausulae, of each phrase according to the rhythmical feet that Cicero acknowledges. We should be aware that this exercise is fraught with various difficulties. There is no clear-cut, unambiguous method, for instance, for the division of ancient texts into phrases, or colons: I concede, therefore, that the divisions below are mere suggestions, although I think that the conspicuous rhythmical conclusions in the majority of the phrases lend credibility to my divisions. 118 Moreover, scansion of prose is a precarious business, for which modern scholars have adopted various approaches. 119 In analyzing the passages cited below, I have applied the following principles: (1) the last syllable of each phrase, whether long or short by nature, is simply marked as long; 120 (2) syllables that are short by nature are

116 Cic. Orat. 230: Sunt etiam qui illo vitio, quod ab Hegesia maxime fluxit, infrgingendis concidendisque numeris in quoddam genus abiectum incidant versiculorum simillimum. Ibid. 226: Nec ullum genus est dicendi aut melius aut fortius quam binis aut ternis ferire verbis, nonnumquam singulis, paulo alias pluribus, inter quae variis clausulis interponit se raro numerosa comprehensio; quam perverse fugiens Hegesias, dum ille quoque imitari Lysiam volt, alterum paene Demosthenem, saltat incidens particulas. 117 Dion. Hal. Dem. 50.11 praises Demosthenes for the varying length of his sentences, alternately using colons and periods. Cf. Demetr. Eloc. 241–245, who, like Cicero, advises to use rhythmical periods sparingly. Cicero mentions Hegesias’ admiration for Lysias as a warning to the self-styled ‘Attic’ orators in Rome, who also preferred Lysias as their model: cf. section 2.2 esp. n. 28–29 above. 118 Cf. Hutchinson (2015) 796–798 on prose rhythm in Appian (second century AD): ‘Phrase is only a shorthand, since a sizeable single word can form a separate entity for these purposes, and a brief beginning to a sentence or part of it can be followed by a break and the main start. (…) Closes that appear rhythmical occur in Appian not just at the end of sentences or half-sentences but throughout the sentence.’ 119 For scansions of Hegesias’ prose, see Norden (1898) 135–137 on F12 and F24 Jacoby, Blass (1905) 19–23 on F5 and F12 Jacoby. Cf. also De Groot (1919a) 128–131, (1919b) 6, (1921) 64, and Skimina (1937) 144–148. 120 Short syllables at phrase-end are considered to be ‘lengthened by quasi-metrical pause’: cf. Hutchinson (2015) 789. It is a widely accepted convention in modern scansion of Greek poetry to mark naturally short

CHAPTER TWO marked as either long or short (i.e., ͞ᴗ), if they precede a juxtaposition of mute and liquid (but they are marked as long, if they precede γμ , γν , δμ or δν );121 and (3) collisions of vowels will be assumed to result in hiatus.122 With these ground rules in mind, we can address the rhythmical phrases in the three longest extant fragments of Hegesias, preserved in Dionysius’ On the Arrangement of Words (F5 Jacoby), Agatharchides’ On the Red Sea (F12 Jacoby) and Strabo’s Geography (F24 Jacoby). 123 The accompanying notes to the former fragment show that the transmission of Dionysius’ citation is flawed in several places.124

F5 ὁ δὲ βασιλεὺς ἔχων τὸ σύνταγμα προηγεῖτο . ͞ᴗ ᴗ ― | ― ― Καί πως ἐβεβούλευτο τῶν πολεμίων τοῖς ἀρίστοις , ᴗ ᴗ ᴗ ― | ― ᴗ | ― ― 3 ἀπαντᾶν εἰσιόντι· ― ― | ― ᴗ | ― ― τοῦτο γὰρ ἔγνωστο , ― ᴗ ᴗ ― | ― ― κρατήσασιν ἑνὸς συνεκβαλεῖν καὶ τὸ πλῆθος . ― ᴗ ― | ― ͞ᴗ | ― ― 6 Ἡ μὲν οὖν ἐλπὶς αὕτη συνέδραμεν εἰς τὸ τολμᾶν,125 ͞ᴗ ᴗ ᴗ | ― ᴗ | ― ― ὥστ ’ Ἀλέξανδρον μηδέποτε κινδυνεῦσαι πρότερον οὕτως . ― ᴗ ᴗ ᴗ | ― ― syllables at verse-end as long, regarding them as (syllaba ) brevis in (elemento ) longo : cf., e.g., Korzeniewski (1968) 28 and West (1982) 35. Note that in ancient discussions of prose rhythm, there was some disagreement about the quantity of the final syllable of the period. Cic. Orat. 217–218, for instance, insists that it is indifferent: ‘For it never makes any difference whether the dactyl ( ― ᴗ ᴗ) or cretic ( ― ᴗ ―) be used last, because even in verse the quantity of the final syllable is a matter of indifference’ ( nihil enim interest dactylus sit extremus an creticus, quia postrema syllaba brevis an longa sit ne in versu quidem refert ). Yet, Arist. Rh. 3.8.6 insists that the period ‘should be broken off by a long syllable’ ( δεῖ τῇ μακρᾷ ἀποκόπτεσθαι ), while Quint. Inst. orat. 9.4.94 notes: ‘I am aware of course that a short is treated as a long in final position, because there seems to be a bit of vacant time accruing to it from what follows; however, when I consult my own ears, I realize that it makes a great difference whether the closing syllable is really long or merely treated as long’ ( neque enim ego ignoro in fine pro longa accipi brevem, quia videtur aliquid vacantis temporis ex eo quod insequitur accedere: aures tamen consulens meas intellego multum referre verene longa sit quae cludit an pro longa ). 121 Cf. Smyth (1920) 35–36 and West (1982) 16. 122 This is in accordance with most Greek discussions of vowel collision in prose: cf. section 4.4 below. Hutchinson (2015) shows that Appian (second century AD) deliberately created hiatus in order to produce more desirable clausulae. On a related note, Nisbet (1990) and Hutchinson (1995) have demonstrated that Cicero uses atque (instead of ac or et ) before a consonant, if the combination yields a more favored clausula. There are only two instances of vowel collision in the clausulae of the texts below: see the notes at lines 20 and 26 of F5 Jacoby. 123 Dion. Hal. Comp. 18.26, Agatharch. Eryhtr. 120.35–42 Müller (= Phot. Bibl. 446b9–15), Str. Geogr. 9.1.16. 124 I have used the edition of Aujac and Lebel (1981). Donadi (2000a) 338–342 compares the variant readings of the two mss F and P., noting that Aujac and Lebel select the readings that yield Hegesias’ favorite clausulae. 125 Usener and Radermacher read συνέδραμεν εἰς τόλμαν , yielding an uncommon rhythm ( ᴗ ͞ᴗ ᴗ ᴗ ― | ― ―).

HERO VERSUS ZERO

126 F has εἰς γόνατα συγκαθίσας , which gives an unusual rhythm ( ― ᴗ ᴗ ᴗ ― ᴗ ᴗ ―) 127 P has πρόσφατος ἐπὶ παλαιαῖς, resulting in a rarely encountered cadence ( ― ᴗ ᴗ ᴗ ᴗ ᴗ | ― ―). 128 F has τολμήσαντος ἀπόνοια , yielding an uncommon ending ( ― ― | ― ᴗ ᴗ ᴗ | ― ―). 129 In poetry, a short final α is often elided: cf. West (1982) 10. This would yield a ditrochee ( ― ᴗ | ― ―). 130 In poetry, τό often blends with the subsequent vowel: cf. West (1982) 13. This results in one of Hegesias’ favorite endings ( ― ᴗ ― | ― ―). 131 F has ζῷον ἕτερον ἁδρόν , which possibly gives a slightly different cadence ( ― ᴗ ᴗ ᴗ ᴗ | ͞ᴗ ―).

CHAPTER TWO

F12 Ὅμοιον πεποίηκας , Ἀλέξανδρε , ― ᴗ ᴗ ― | ― ― Θήβας κατασκάψας , ― ᴗ ― | ― ― 3 ὡς ἂν εἰ ὁ Ζεὺς ἐκ τῆς κατ ’ οὐρανὸν μερίδος ἐκβάλοι τὴν σελήνην· ― ᴗ ― | ― ᴗ | ― ― ὑπολείπομαι γὰρ τὸν ἥλιον ταῖς Ἀθήναις . ― ᴗ ― | ― ᴗ | ― ― Δύο γὰρ αὖται πόλεις τῆς Ἑλλάδος ἦσαν ὄψεις . ― ᴗ ᴗ | ― ᴗ | ― ― 6 Διὸ καὶ περὶ τῆς ἑτέρας ἀγωνιῶ νῦν· ― ᴗ | ― ᴗ | ― ― ὁ μὲν γὰρ εἷς αὐτῶν ὀφθαλμὸς ἡ Θηβαίων ἐκκέκοπται πόλις . ― ᴗ ― | ― ᴗ ―

I do not wish to make any claims about Hegesias’ stylistic practices; I simply submit that one can see, on the basis of the scansions above, why Cicero might consider Hegasias’ prose to be a string of rhythmical chunks. Indeed, one who follows Cicero’s rules for dividing rhythmical feet may recognize two types of clausula in Hegesias. First, between eleven and fourteen phrases end on a so-called ditrochee ( ― ᴗ | ― ―), covering about a quarter (between 20,4% and 25,9%) of the fifty-four quoted phrases.132 In his Orator , Cicero connects this cadence to Asian oratory, praising its devastating effect, but warning that it must be used sparingly: ‘This ending ought not to be used too frequently, for first it is recognized as rhythm, next it wearies, and when it begins to seem an easy trick, it is despised.’ 133 In the samples above, final ditrochees are often, in three to six cases, preceded by a cretic (― ᴗ ― | ― ᴗ | ― ―).134

132 Hegesias’ fondness of ditrochaic clausulae is well attested: cf. Norden (1898) 135–137, Blass (1905) 20, De Groot (1919a) 128. Note that Blass, dividing F5 Jacoby into thirty-nine rhythmical phrases, argues that twenty- one end on a ditrochee, including ten cases in which the first syllable is allegedly resolved ( ᴗ ᴗ ᴗ | ― ―). As we will see in the present paragraph, this variant cadence can often be treated as paean + spondee ( ― ᴗ ᴗ ᴗ | ― ―), a clausula that Cic. Orat. 232 seems to recommend (cf. section 4.4 n. 80 below). 133 Cic. Orat. 215: Sed id crebrius fieri non oportet; primum enim numerus agnoscitur, deinde satiat, postea cognita facilitate contemnitur. On the ditrochee as typical for Asian oratory, see ibid. 212: ‘Asia prefers the one called ditrochee’ ( unum est secuta Asia macime qui dichoreus vocatur ). Cf. section 2.3.3 n. 90 above on Cicero’s terminology ( choreus ) for the trochee, and section 4.4 below on Cicero’s praise of one-word ditrochees, citing two examples from C. Carbo, viz., persolutas and comprobavit . 134 Cf. De Groot (1919a) 128 and (1921) 64.

HERO VERSUS ZERO

Secondly, Hegesias also seems to have liked clausulae consisting of paean + spondee (either ― ᴗ ᴗ ᴗ | ― ― or ᴗ ᴗ ᴗ ― | ― ― ) and cretic + spondee ( ― ᴗ ― | ― ―): the former ending occurs between ten and thirteen times, the latter is attested between five and eight times. 135 Considering that Cicero explicitly declares paeans and cretics to be interchangeable, we may treat these clausulae as a single group, occurring in approximately a third of the quoted phrases (between 27,8% and 38,9%). 136 To be brief, the three most substantial fragments of Hegesias show that Cicero could readily divide them up into little verse-like phrases whose conclusions exhibit little variation: in majority (between 57,4% and 70,4%), they end either on a ditrochee or on a combination of penultimate cretic/paean and final spondee. That Cicero was aware of Hegesias’ preference for these clausulae becomes obvious in one of his letters to Atticus (May 46 BC). The Roman rhetorician provides his friend with a self-written mock-specimen of Hegesias’ style: ‘Here you have a sample of Hegesias’ type of style ( genus Hegesiae ), which Varro commends.’ 137 Like Hegesias’ own texts, Cicero’s parody can be cut up into particulae and versiculi :138

De Caelio vide quaeso, ― ― ᴗ | ― ᴗ ― | ― ― ne lacuna sit in auro. ― ᴗ | ― ᴗ ᴗ ᴗ | ― ― 3 Ego ista non novi, ᴗ | ― ᴗ ― | ― ― sed certe in collubo est ― ― ― | ― ᴗ ― detrimenti satis. ― ― ― | ― ᴗ ― 6 Huc aurum si accedit ― ― | ― ― | ― ―

135 The sequence cretic + spondee is generally recognized as one of Hegesias’ favorite cadences: see Norden (1898) 135–137, De Groot (1921) 64. Blass (1905) notes that iambus + spondee ( ᴗ ― | ― ―) is common, occurring in nine out of the thirty-nine clauses that he recognizes in F5 Jacoby. For the preponderance of paean + spondee, see De Groot (1919a) 128, who lists this clausula together with the ditrochee as Hegesias’ ‘most favored’ ending. Curiously, Cicero was heavily criticized for ending his sentences frequently on the words esse videatur , that is, on paean + spondee, one of Hegesias’ favorite rhythms: cf. section 4.2 n. 36 below. 136 Cic. De or. 3.193, Orat. 215. 137 Cic. Att. 12.6.1: Habes Hegesiae genus, quod Varro laudat. Cf. section 2.1 n. 6 above. 138 I assume collisions of vowels to result in elision or prodelision: thus, I read cert(e) in, s(i) accedit , and collubo (e)st. After all, Cic. Orat. 150–152 himself claims that the Latin language, unlike Greek, demands its speakers to ‘run vowels together’ ( vocalis coniungere ) and not to ‘make a pause between vowels’ ( distrahere voces ): cf. section 4.5 below. Shackleton Bailey (1999) translates the parody as follows: ‘Please see about Caelius and see that there is no gap in the gold. I know nothing about these matters. But surely there’s enough lost on the exchange. If the gold comes up on top of that—but why talk? You’ll see to it.’

CHAPTER TWO

—sed quod loquor? ― | ― ᴗ ― Tu videbis. ― ᴗ | ― ―

In the first three consecutive clauses, we recognize the closing rhythm that we have encountered most often in Hegesias’ largest fragments, that is, cretic or paean + spondee. Next, Cicero composes two parallel phrases that follow exactly the same rhythmical pattern, consisting of a molossus ( ― ― ―) and a cretic ( ― ᴗ ―), which, to my knowledge, is not attested in Hegesias, but which was certainly recognized as a rhythmical cadence. 139 The final phrase in this quasi-Hegesian passage appropriately ends on a ditrochee, which the Hellenistic author, as we have seen, seems to employ frequently, and which, according to Cicero, is Asia’s favorite clausula. All in all, Cicero takes exception to Hegesias’ overly rhythmical style: the man from Magnesia did not apply his rhythms with moderation, but he used them obtrusively throughout his discourse. 140 An interesting parallel to Cicero’s position can be found in Aelius Theon: ‘You should be careful in the arrangement of your words, and learn how to avoid the mistakes of bad arrangement, especially a style that is in meter ( ἔμμετος ) and in rhythm (ἔρρυθμος ), such as most of the works of the orator Hegesias and the so-called Asian orators exhibit.’ 141 Thus, both Cicero and Theon accuse him of committing the very error that Demosthenes, according to Dionysius, expertly avoided: while the Attic orator could be praised for making his style appear ‘rhythmical’ ( εὔρυθμος ) and ‘metrical’ ( εὔμετρος ), without being ‘in rhythm’ and ‘in meter’ (section 2.3.3), Hegesias was seen as a slave to his own monotone, almost singsong rhythmical format. His alleged obsession with such artistic flourish makes him Cicero’s ultimate archetype of Asianism gone wrong: although the rhetorician is, as we have seen, not unequivocally negative about Asian oratory (sections 1.6

139 See e.g. Zieli ński (1904) 223–225, Berry (1996) 58 and Von Albrecht (2003) 111 on this pattern in Cicero; cf. Hutchinson (2015) 789 on molossus + cretic in Greek imperial prose. 140 Cf. Cic. Orat. 220: ‘It makes a vast difference whether prose is rhythmical, that is resembling definite rhythms, or is composed entirely of definite rhythms; if the latter occurs, it is an intolerable fault; if the former does not occur, the style is disordered, unpolished and vague’ ( multum interest utrum numerosa sit, id est similis numerorum, an plane e numeris constet oratio; alterum si fit, intolerabile vitium est, alterum nisi fit, dissipata et inculta et fluens est oratio ). Cf. the next note below. 141 Theon Prog. p. 71.7–11 Spengel: Ἐπιμελητέον δὲ καὶ τῆς συνθέσεως τῶν ὀνομάτων , πάντα διδάσκοντα ἐξ ὧν διαφεύξομαι τὸ κακῶς συντιθέναι , καὶ μάλιστα δὲ τὴν ἔμμετρον καὶ ἔρρυθμον λέξιν , ὡς τὰ πολλὰ τῶν Ἡγησίου τοῦ ῥήτορος καὶ τῶν Ἀσιανῶν καλουμένων ῥητόρων . For the distinction between ‘rhythmical’ and ‘in rhythm’, which goes back to Arist. Rh. 3.8.3, see section 2.3.3 n. 89 above.

HERO VERSUS ZERO and 5.6.1), he does submit that it conveys ‘a rather small amount of authority’ ( auctoritatis parum ), as it is better suited for youth than for old age. 142 This holds good, a fortiori, for Hegesias: ‘Where will you find anything so broken (fractum ), so minced ( minutum ), anything so immature as that balance (concinnitate puerile ) that he cultivated?’ 143 According to Cicero, then, Hegesias’ style is ill-suited for serious oratory: this point was already made, as we have seen, by Agatharchides (section 2.4.1), and it will be made again, as we will see, by Dionysius (section 2.4.3). Cicero refers twice to Hegesias’ prose as a clearly demarcated ‘type of style’ ( genus ) on its own. Apparently, the three existing ‘types of style’ ( genera dicendi ) fall short in describing Hegesias’ peculiar style. What is more, Hegesias is not even mentioned in Cicero’s discussion of the two ‘types of Asian style’ (genera Asiaticae dictionis ) in his Brutus : the ‘pithy and clever variety’ (genus sententiosum et argutum ) is linked to the historian Timaeus and the brothers Hierocles and Menecles of Alabanda, whereas the ‘swift and impetuous variety’ ( genus volucre atque incitatum ) is connected with Aeschylus of Cnidus and Aeschines of Miletus.144 Why does Cicero omit Hegesias, in his view the founding father of Asianism, in this typology of Asian oratory? While Cicero is generally ambivalent toward Asian oratory (sections 1.6 and 5.6.1), he considers the Magnesian’s style too bizarre for any classification, worthy of nothing but scorn and ridicule. After all, unlike Demosthenes’ near-perfect Attic prose, it conspicuously lacks variation, it is unsuitable for conveying serious thoughts, and it displays a staggering incompetence in the art of rhythmical word arrangement.

142 Cic. Brut. 326–327, discussing Hortensius. Cf. section 5.6.1 below on Hortensius’ Asianism and section 1.6 above on Cicero’s ambivalent approach to Asianism. 143 Cic. Brut. 286: At quid est tam fractum, tam minutum, tam in ipsa, quam tamen consequitur, concinnitate puerile. On Hegesias’ puerility, cf. Agatharch. Erythr. p. 120.24 Müller (= Phot. Bibl. 446a40), cited in section 2.4.1 n.109 above : αἴτημα μειρακιῶδες . In Cicero, the term concinnitas most often refers to the use of various figures of speech to create some form of balance in the sentence: see Cic. Orat. 149, 164. Cf. Rhet. Her. 4.18, Sen. Ep. 115.2, Quint. Inst. orat. 9.1.34, 9.3.91. Hegesias’ attention to such figures is clearly exhibited in the samples quoted by Agatharchides (section 2.4.1). Latin translations of some of Hegesias’ figures of speech survive through Rutilius Lupus’ excerpt (= Hegesias F27–29 Jacoby) of the treatise On Figures of Speech by the Greek rhetor Gorgias (first century BC): see n. 94 above. 144 Cic. Brut. 325–327. The description of the former variant comes closest in describing Hegesias’ style: it is less characterized ‘by weight of thought than by the charm of balance’ ( non tam gravibus et severis quam concinnis et venustis ). The latter variant, which ‘lacked elaborate balance of sentence’ ( ornata sententiarum concinnitas non erat ), does not approximate the prose of Hegesias, who, according to Cicero, was always hunting for symmetry in his sentence: cf. the previous note. Cf. Kennedy (1994) 96; according to Leeman (1963) 95, the two types of Asianism were combined by Hegesias.

CHAPTER TWO

2.4.3 Dionysius on Hegesias’ Style Like Cicero, Dionysius is outraged by Hegesias’ word arrangement. Unlike Cicero, however, the Greek critic does not go into the author’s staccato sentences with their obvious rhythmical cadences. In On the Arrangement of Words , Dionysius focuses on two other aspects of Hegesias’ composition: he criticizes (1) his inability to adapt his rhythm to the seriousness of the topic at hand, and (2) his affected, far-fetched word order. In order to illustrate the former point, Dionysius compares the (in his view) noble rhythms of Thucydides, Demosthenes, Plato and Homer to the Magnesian’s supposedly ignoble rhythms; in order to elucidate the latter point, he recasts a passage from Herodotus into the styles of Thucydides and Hegesias respectively, merely by rearranging the original words. First, Dionysius stresses that the grandeur or banality of any subject matter should be reflected in the arrangement of the words. In his view, the power of arrangement in literature can be compared to the role that the goddess Athena plays in the Odyssey: ‘It seems to me that one would not be wrong to compare word arrangement to Athena in Homer, for she used to make the same Odysseus appear in different forms at different times (…). So also does arrangement take the same words and make the ideas that they convey appear misshapen, beggarly and mean, and at other times sublime, rich and beautiful.’145 According to Dionysius, one important way of magically changing the appearance of poetry or prose is through rhythm: he distinguishes between rhythmical feet that allow an author to make his text appear ‘august’ ( σεμνός ), ‘elevated’ ( ὑψηλός ) and ‘manly’ ( ἀνδρώδης ), and others that make the words sound ‘ignoble’ (ἀγεννής ), ‘low’ ( ταπεινός ) and ‘feminine’ ( θῆλυς ). As we have seen, Dionysius ranks rhythmical feet that consist preponderantly of long syllables in the former category, while he places rhythms that are dominated by short syllables in the latter group (section 2.3.3). 146 After cataloguing the twelve disyllabic and trisyllabic rhythmical feet

145 Dion. Hal. Comp. 4.12: Καί μοι δοκεῖ τις οὐκ ἂν ἁμαρτάνειν εἰκάσας αὐτὴν τῇ Ὁμηρικῇ Ἀθηνᾷ· ἐκείνη τε γὰρ τὸν Ὀδυσσέα τὸν αὐτὸν ὄντα ἄλλοτε ἀλλοῖον ἐποίει φαίνεσθαι , (...). Αὕτη τε τὰ αὐτὰ λαμβάνουσα ὀνόματα , τοτὲ μὲν ἄμορφα καὶ πτωχὰ καὶ ταπεινὰ ποιεῖ φαίνεσθαι τὰ νοήματα , τοτὲ δ’ ὑψηλὰ καὶ πλούσια καὶ καλά . Sadée was probably right in deleting the words [καὶ ἁδρά ] after πλούσια , as they destroy the symmetry of the three, chiastically organized, antitheses between high and low ( ὑψηλά , ταπεινά ), rich and poor ( πλούσια , πτωχά ), and beautiful and ugly (καλά , ἄμορφα ): see Aujac and Lebel (1981) 203. In addition, the adjective ἁδρός does not belong to Dionysius’ terminology for the elevated style of prose: cf. section 3.2 below. 146 Dion. Hal. Comp. 17 ranks twelve disyllabic and trisyllabic feet according to these two categories. In the ‘noble’ category are ranked the spondee ( ― ―), molossus ( ― ― ―), anapest ( ᴗ ᴗ ―), dactyl ( ― ᴗ ᴗ), bacchius (― ― ᴗ) and hypobacchius ( ᴗ ― ―); the ‘ignoble’ category includes the pyrrhic ( ᴗ ᴗ), choree ( ᴗ ᴗ ᴗ) trochee (― ᴗ), and amphibrach ( ᴗ ― ᴗ). Cf. section 2.3.3 n. 90 above. Note that Dion. Hal. Comp. 17.5, 17.13 calls both

HERO VERSUS ZERO in this fashion, Dionysius turns to five passages that treat grand topics full of pathos, showing how rhythm may reinforce or, in Hegesias’ case, destroy the dignity of the substance. After scanning samples from Thucydides’ account of Pericles’ funeral oration, the encomium of the fallen sons of Athens in Plato’s Menexenus , and Demosthenes’ vehement apology for his politics in On the Crown , Dionysius remarks: ‘So what was there to prevent the arrangement from being beautiful in a passage which contains no pyrrhic foot ( ᴗ ᴗ), nor any iambus ( ᴗ ―), amphibrach ( ᴗ ― ᴗ), choree ( ᴗ ᴗ ᴗ) or trochee ( ― ᴗ)? Now I do not say that none of these authors ever use the more ignoble rhythms also; for they do. But they have concealed them well, dispersing them and interweaving the inferior ones with the better.’ 147 Indeed, Dionysius, rather arbitrarily, only recognizes rhythms in these passages that he previously marked as noble, viz., the molossus ( ― ― ―), the spondee ( ― ―), the bacchius (― ― ᴗ), the hypobacchius ( ᴗ ― ―), the cretic ( ― ᴗ ―), the dactyl ( ― ᴗ ᴗ) and the anapest (ᴗ ᴗ ―).148 Hegesias’ rhythms, conversely, display the opposite tendency, according to Dionysius: ‘Those authors who have not taken any precaution concerning this part of their craft have produced writings which are either mean ( ταπεινός ) or enfeebled (κατακεκλασμένος ), or have some other deformity or disfigurement (ἄλλη τις αἰσχύνη καὶ ἀμορφία ). The first, middlemost and last in this is that sophist from Magnesia, Hegesias.’ 149

the cretic ( ― ᴗ ―) and the iambus ( ᴗ ―) ‘not undignified’ ( οὐκ ἀγεννής ): yet, as we will see in the next paragraph, the critic tends to associate the former foot with nobility and the latter with a lack thereof. The paean (― ᴗ ᴗ ᴗ or ᴗ ᴗ ᴗ ―), favored by Arist. Rh. 3.8.4–6, falls outside the scope of Dionysius’ division, but see n. 153 below. Cf. section 4.4 below on rhythmical feet in Dionysius’ theory of three styles of word arrangement. 147 Dion. Hal. Comp. 18.20: Τί οὖν ἐκώλυε καλὴν ἁρμονίαν εἶναι λέξεως , ἐν ᾗ μήτε πυρρίχιός ἐστι ποὺς μήτε ἰαμβικὸς μήτε ἀμφίβραχυς μήτε τῶν χορείων ἢ τῶν τροχαίων μηδείς; Καὶ οὐ λέγω τοῦτο , ὅτι τῶν ἀνδρῶν ἐκείνων ἕκαστος οὐ κέχρηταί ποτε καὶ τοῖς ἀγεννεστέροις ῥυθμοῖς ( κέχρηται γάρ ), ἀλλ ’ εὖ συγκεκρύφασιν αὐτοὺς καὶ συνυφάγκασι διαλαβόντες τοῖς κρείττοσι τοὺς χείρονας . Dionysius discusses the rhythm in Thuc. 2.35 (Pericles’ funerary speech), Pl. Menex. 236d, Dem. De cor. 1. These three speeches are among Cicero’s best-loved orations: cf. Brut. 27, Orat. 29 on Pericles, ibid. 151 on Pl. Menex. and ibid. 133 on Dem. De cor. , Cicero’s absolute favorite. Cf. section 2.2 n. 20 above for the ancient appreciation of De cor. , esp. its patriotism. 148 De Jonge (2008) 340–347 addresses the awkwardness of Dionysius’ scansions: ‘Dionysius’ analysis of rhythmical prose is indeed problematic: his divisions of clauses into metrical feet seem to be rather arbitrary, sometimes even inconsistent.’ The critic provides two incompatible scansions of Dem. De cor. 1, which, according to De Jonge, can be explained by considering the ‘local contexts’: in Comp. 18.15–20, the critic aims to demonstrate the dignity of Demosthenes’ rhythm, but in Comp. 25.26–28, he attempts to show that Demosthenes’ prose can be divided into near-perfect metrical verses. 149 Dion. Hal. Comp. 18.21–22: Οἷς δὲ μὴ ἐγένετο πρόνοια τούτου τοῦ μέρους , οἳ μὲν ταπεινάς , οἳ δὲ κατακλασμένας , οἳ δ’ ἄλλην τινὰ αἰσχύνην ἢ ἀμορφίαν ἐχούσας ἐξήνεγκαν τὰς γραφάς . Ὧν ἐστι πρῶτός τε καὶ

CHAPTER TWO

In these unmistakable terms, Dionysius introduces the Asian author as his ultimate archetype of bad rhythm. In accordance with Agatharchides’ judgment, Dionysius considers the incongruity of style and substance in Hegesias’ prose a clear sign of the author’s madness: ‘Was he so insensitive and dense that he could not envisage which are the ignoble or the noble rhythms? Or was he so bedeviled and mentally deranged (θεοβλάβεια καὶ διαφθορὰ τῶν φρενῶν) that he still chose the worse, though he knew the better? I am inclined to believe the latter: for it is a characteristic of ignorance (ἄγνοια ) that it often lands on its feet; willfulness (πρόνοια ) never does.’ 150 Το demonstrate Hegesias’ deliberate corruption, Dionysius quotes a long sample describing how Alexander ties a Babylonian king to his chariot and drags him around on his encampment in Gaza: we have already seen that this fragment, full of Asian stereotypes, confirms Cicero’s view that the Magnesian’s style is permeated with jerky verse- like phrases (section 2.4.2). 151 Dionysius, however, compares the passage to the equally grave episode in Homer’s Iliad , in which Achilles attaches the body of Hector to his wagon: ‘What, then, is the cause of the nobility of these lines, and of the miserable inadequacy of the other drivel (τῶν φλυαρημάτων ταπεινότης )? The main cause, if not the only one, is the difference in the rhythms. In the passage of Homer, there is not a single undignified or undistinguished line, whereas in that from Hegesias not a single sentence fails to give offense.’ 152

τελευταῖος καὶ μέσος ὁ Μάγνης σοφιστὴς Ἡγησίας . Ὧν ἐστι πρῶτός τε καὶ τελευταῖος καὶ μέσος ὁ Μάγνης σοφιστὴς Ἡγησίας . The verb ‘to break short’ ( κατακλάω ) does not seem to refer, at least not primarily, to the jerkiness of Hegesias’ style, to which Cicero took exception (section 2.4.2), but rather to the effeminate character of his prose: for this connotation of κατακεκλασμένος , see e.g. com. adesp. 339.2 and Luc. Symp. 18. For the effeminacy of Hegesias, see Dion. Hal. Comp. 4.11 ( σχῆμα μαλθακόν , cf. n. 159 below), 18.28 ( ὑπὸ γυναικῶν ἢ κατεαγότων ἀνθρώπων λέγοιτ ’ ἄν, cf. n. 7 above) ; see section 5.3 below on the femininity associated with Asia. 150 Dion. Hal. Comp. 18.22: Πότερα τοσαύτη περὶ αὐτὸν ἦν ἀναισθησία καὶ παχύτης ὥστε μὴ συνορᾶν οἵτινές εἰσιν εὐγενεῖς ἢ ἀγεννεῖς ῥυθμοί , ἢ τοσαύτη θεοβλάβεια καὶ διαφθορὰ τῶν φρενῶν ὥστ ’ εἰδότα τοὺς κρείττους ἔπειτα αἱρεῖσθαι τοὺς χείρονας , ὃ καὶ μᾶλλον πείθομαι . Ἀγνοίας μὲν γάρ ἐστι καὶ τὸ κατορθοῦν πολλαχῇ, προνοίας δὲ τὸ μηδέποτε . Cf. Agatharch. Erythr. 121.36 Müller (= Phot. Bibl. 447a14): μανία . 151 Dion. Hal. Comp. 18.26 (= Hegesias F5 Jacoby). See section 2.2 above for the Asian flavor of the passage; see section 2.4.2 above for the complete text with an analysis of its minute phrases and rhythmical cadences. Spina (1989) compares this passage to other sources on Alexander’s siege of Gaza (332 BC), i.e., Curt. 4.6.7–30, Arr. Anab. 25.4–27.21. Cf. section 2.4.1 n. 98 above on the identity of the work from which Dionysius quotes the passage: Dion. Hal. Comp. 18.24 refers to it as a ‘passage from his history’ ( λέξις ἐξ ἱστορίας ). 152 Dion. Hal. Comp. 18.29: Τί οὖν αἴτιον ἐκείνων μὲν τῶν ποιημάτων τῆς εὐγενείας , τούτων δὲ τῶν φλυαρημάτων τῆς ταπεινότητος; Ἡ τῶν ῥυθμῶν διαφθορὰ μάλιστα , καὶ εἰ μὴ μόνη . Ἐν ἐκείνοις μὲν γὰρ οὐδεὶς στίχος ἄσεμνος , ἐνταῦθα δὲ οὐδεμία περίοδος ἥτις οὐ λυπήσει . Dionysius quotes Hom. Il. 22.395–411.

HERO VERSUS ZERO

The comparison between Homer and Hegesias is somewhat forced: although the passages are thematically linked, it seems a bit unfair to juxtapose their rhythms, as Homer’s repertoire is necessarily limited to the spondee ( ― ―) and the dactyl ( ― ᴗ ᴗ), which, in Dionysius’ view, are among the noblest of feet. To make things even more complicated, the critic does not scan Hegesias’ sentences; as he is not at all consistent in his scansions of other passages, there is no way of knowing how he would divide the text into rhythmical feet. We do know, however, that Dionysius considers sequences of short syllables undignified. In the quoted passages from Thucydides and Plato, there is not a single string of three or more short syllables, while Demosthenes, as we have seen, generally avoided such rhythmical patterns (section 2.3.3). 153 Hegesias’ large Alexander fragment, however, already begins with four consecutive short syllables: ὁ δὲ βασιλεύς (ᴗ ᴗ ᴗ ᴗ ―). In addition, it contains several word combinations that suggest a propensity for swift rhythms, as can be gleaned from the complete quotation included in the previous section: take for instance τὸ ξίφος ἐνέγκαντος ἐπὶ τὰ πτερύγια ( ― ᴗ ᴗ ᴗ ― ― ᴗ ᴗ ᴗ ͞ᴗ ᴗ ᴗ ᴗ ᴗ), τὸ συνάγαγον (ᴗ ᴗ ᴗ ᴗ ᴗ), and γαστρὸς ἐνέφαινε Βαβυλώνιον ζῷον ἕτερον ἀνδρός ( ― ᴗ ᴗ ᴗ ― ᴗ ᴗ ᴗ ― ᴗ ― ― ᴗ ᴗ ᴗ ᴗ ― ᴗ). There is no doubt that Dionysius would write such rhythmical patterns off as undignified, unmanly and unsuitable for serious discourse. 154 For Dionysius’ second objection against Hegesias’ word arrangement, I will turn to an intriguing self-written parody of the Magnesian’s style. The critic compares the ‘Hegesian style of arrangement’ (Ἡγησιακὸν σχῆμα τῆς συνθέσεως ) to the practice of his fellow historians Herodotus and Thucydides. Remarkably, Dionysius does not illustrate his point by selecting samples from all three authors, but he rather uses a single passage from Herodotus, which he then rewrites once in the style of Thucydides, and again in the style of Hegesias. This is a rather creative application of the method of ‘rewriting’ (μετάθεσις ), which Dionysius

153 In the quoted passage from Dem. De cor. 1, Dionysius points out two places, where there are three consecutive short syllables, that is, in ἐγὼ διατελῶ τῇ τε πόλει (ᴗ ― ᴗ ᴗ ᴗ ― ― ᴗ ᴗ ᴗ): here, the critic does not censure Demosthenes for twice using the ignoble choree, but he rather recognizes two subsequent paeans, the use of which, as we have seen, was propagated by Arist. Rh. 3.8.4–6, but which Dionysius himself only mentions here, vindicating his Attic hero. The scansion of πόλει (ᴗ ᴗ) is surprising: De Jonge (2008) 343 suggests that Dionysius’ analysis ‘may reflect certain changes in the perception of the quantities of syllables’. Similar remarkable scansions in this passage are εὔνοιαν ( ― ― ͞ᴗ) and τόν ( ―). Yet, in the scansion of the same passage in Comp. 25.26–28, the quantities of the syllables do follow the metrical rules: cf. n. 148 above. 154 In fact, both in Greek and Latin stylistic theory, long syllables are often associated with dignity, and short syllables with a lack thereof. Cf. section 2.3.3 n. 90 above on the connection between short syllables and dance. See also section 4.4 below on the aptness of long syllables for producing rough word arrangements.

CHAPTER TWO frequently uses to highlight the virtues or vices of a specific text, or, as in the present case, to compare alternative styles.155 The method seems to have been an integral part of the shared Greek and Roman critical toolkit: it can be found in Philodemus, Demetrius, Longinus, Hermogenes and Cicero. 156 The Roman scholar posits that metathesis can demonstrate the characteristics of a certain composition, ‘when the order of the words is slightly changed, though the words are the same and the thought is the same’ ( ordine verborum paululum commutato, isdem tamen verbis, stante sententia ). Likewise, Dionysius proposes that the quality of a given arrangement can be assessed, ‘when the words are retained but their order is changed’ ( μενόντων μὲν τῶν ὀνομάτων , ἀλλαττομένης δὲ τῆς συνθέσεως ). 157 Dionysius adds brief stylistic commentaries to his two rewritings as well as to the original text. 158 He describes the Herodotean original as ‘relaxed’ ( ὑπαγωγικός ) and ‘historical’ ( ἱστορικός ), while his Thucydidean rearrangement is supposed to be ‘direct’ (ὀρθός ) and ‘engaging’ (ἐναγώνιος ). The Hegesian version, finally, is advertised, in familiar

155 Bonner (1939) 62–70, 102–104 has shown that Dionysius uses the method of ‘metathesis’, which Bonner translates as ‘recasting’, more often in his later oeuvre than in his earlier works: it is used sparingly in the early- period works Isoc. and Is. , it is ‘increased and extended’ in the middle-period treatises Dem. and Comp. , and it is employed even more frequently in the late-period works Thuc. and Amm. II. Cf. Damon (1991) 51–52. For a systematic analysis of the critic’s versatile application of metathesis, see De Jonge (2005) and, in a marginally adapted form, De Jonge (2008) 367–390: De Jonge recognizes three categories of metatheses, (1) ‘metatheses correcting alleged faults of the original’, (2) ‘metatheses bringing out virtues of the original’, and (3) ‘metatheses illustrating alternative compositions or particularities’, which include conversions of the Ionic dialect, comparisons of arrangement styles (incl. Comp. 4.7–11), and illustrations of the poetical character of prose. 156 For the role of metathesis as an evaluative tool in the ancient critical tradition, see Spina (2004a), focusing on various rewritings of Her. 1.8–14 on Gyges and Candaules, Spina (2004b), offering an introduction to grammatical, rhetorical and stylistic rewritings in Antiquity, and Grimaldi (2004), who explores examples from Demetrius, Dionysius and Longinus. As De Jonge (2005) 464 shows, the practice of rewriting texts can already by found in Plato and Aristotle. Demetrius is our most generous source for metathesis: see Janko (2000) 227 n. 2 for a full list. Philodemus and his opponents were involved in a polemic about the possibility of metathesis: see Armstrong (1995a) and Oberhelman and Armstrong (1995). For instances of rewriting in the other authors mentioned, see e.g. Cic. Orat. 214–215, 232–233; and Long. Subl. 39.4, 40.2–3. 157 Cic. Orat. 233, Dion. Hal. Comp. 4.7. See also Cic. Orat. 81 ( verbis mutatis, manente sententia ), Dion. Hal. Comp. 4.5 ( τῆς μὲν ἐκλογῆς τῶν ὀνομάτων τῆς αὐτῆς μενούσης , τῆς δὲ συνθέσεως μόνης μεταπεσούσης ). Nassal (1910) 27 points out the remarkable similarities between Cicero’s and Dionysius’ articulations of this idea. 158 Dion. Hal. Comp. 4.7–11. The original is Hdt. 1.6.1, which Dionysius quotes not in the original Ionic, but rather in the Attic dialect: cf. section 1.5 n. 82 above. There exists a third metathesis of the first three clauses of this passage, in Hermog. Id. 1.3 Rabe: the author argues that it would harm the clarity of the sentence, if Herodotus had opened it with a participle clause ( Κροίσου ὄντος ) instead of a nominative ( Κροῖσος ἦν).

HERO VERSUS ZERO anti-Hegesian catchwords, as ‘precious’ ( μικρόκομψος ), ‘ignoble’ ( ἀγεννής ) and ‘effeminate’ (μαλθακός ). 159 In what way do the three compositions of the same words differ? And what makes Hegesias’ arrangement inferior to the other two versions?

In his Thucydidean recasting, Dionysius distributes the information more systematically and straightforwardly than in Herodotus’ original version. He starts off each of the three initial clauses by clearly demarcating its topic, discussing, in logical order, first Croesus’ family (υἱὸς μέν ), secondly his tribe ( γένος δέ ), and thirdly the extent of his realm ( τύραννος δέ ). His older countryman from Halicarnassus, by contrast, does not arrange these clauses in the same well-planned fashion: he starts with Croesus’ roots, highlighting the adjective ‘Lydian’ (Λυδὸς μέν ) instead of the noun ‘tribe’, while it is only in the subsequent clauses that he

159 Cf. Agatharch. Erythr. p. 120.1 Müller (= Phot. Bibl. 446a19): κομψότης . For Hegesias’ alleged femininity, see n. 149 above. See also the present section above on Dionysius’ criticism of Hegesias’ ignoble rhythm. 160 ‘Croesus was a Lydian by birth, | son of Alyattes, | king of the nations on this side of the river Halys, | which flows from the south | between Syria and Paphlagonia, | and drains to the north | into the sea, which is called Euxine.’ I have added vertical lines between the phrases to facilitate the comparison of the three passages. 161 Note that Dionysius, despite his promise at Comp. 4.7 to use the same words as in the original, replaces παῖς with υἱός , and, in his Thucydidean rearrangement, he swaps ἐξίησι for ἐκδίδωσι . Concerning the former substitution, De Jonge (2005) 477 n. 47 (= De Jonge (2008) 387 n. 69) suggests that Dionysius ‘has observed that παῖς is more common in Herodotus, whereas Thucydides and Hegesias would rather use υἱός ’, but ‘it may also be explained by the fact that the latter word is more familiar in later Greek’.

CHAPTER TWO seems to adopt a more orderly approach to Croesus’ other personalia, referring to his family (παῖς δ’) and his kingdom ( τύραννος δ’) respectively.162 In the remainder of the sentence, too, Herodotus’ text is more ‘relaxed’ ( ὑπαγωγικός ) than Dionysius’ Thucydidean rewriting: the nations under Croesus’ rule, for instance, are gradually introduced in the original ( ἐθνῶν τῶν etc.), while the restyled passage exhibits a more deliberate presentation ( τῶν ... ἐθνῶν). Finally, Dionysius has also recast Herodotus’ seemingly offhand discussion of the watershed of the Halys into a more careful exposition, which neatly follows the river’s course from the south ( ἀπὸ μεσημβρίας ) to the north ( πρὸς βορέαν ἄνεμον ).163 According to Dionysius, Herodotus’ relaxed style is typical for historiographical writing ( ἱστορικός ), while Thucydides’ systematic presentation is suited for ‘performative oratory’ ( ἐναγώνιοι λόγοι ), as it is straightforward, to-the-point and, in a word, ‘engaging’ (ἐναγώνιος ).164 The imitation of Hegesias’ style, conversely, is designed to appear unsuitable for any kind of serious discourse. To begin with, the focus in each of the first three clauses is haphazard: the first clause emphasizes, in the genitive case, the name of Croesus’ father (Ἀλυάττου μέν ), while the subsequent clauses underline his ‘tribe’ ( γένος δέ ) and the ‘peoples on this side of the river Halys’ ( τῶν δ’ etc.). In addition, the word ‘king’ ( τύραννος ) is weirdly

162 Cf. Weil (1844) 130 on Dionysius’ Thucydidean recasting of the first three clauses: ‘Ces trois petites phrases répondent aux questions: quel est le père, quelle est la patrie, quel est l’empire de Crésus? υἱός , γένος , τύραννος , voilà les trois points de départ, les trois cadres à remplir.’ De Jonge (2005) 477 (= De Jonge (2008) 387) notes that the Thucydidean version deals with Croesus’ personalia ‘in outward expansion’. Next, De Jonge mentions some features of rough word arrangement (cf. section 4.4 below) in the Thucydidean sentence, viz., hiatus, clashes of consonants, long hyperbaton (between ὅς and ἐκδίδωσι ), and the smaller number of connectives (deleted τε ). I do not accept De Jonge’s view that ‘the displacement of Λυδός breaks the parallelism between Λυδός , παῖς ( υἱός ) and τύραννος , and creates anastrophe ’: on the contrary, the change enhances the parallelism. 163 Cf. De Jonge (2005) 477 (= De Jonge (2008) 387): ‘The relative clause flows together with the river Halys’. 164 For historiography and performative oratory in Dionysius’ rewritings, see Bottai (1999b) 145. Ooms and De Jonge (2013) 100–102 note that the term ἐναγώνιος in Dionysius usually means ‘suited for performative oratory’, that is, apt for forensic and deliberative speeches, which he calls ἐναγώνιοι λόγοι : cf. section 2.3.3 n. 78 above. In non-oratorical contexts, the term ἐναγώνιος can express the idea that ‘all participants in the communicative situation are directly concerned with one another and with the subject that is at stake’. Hence, the stylistic attribute ‘engaging’ ( ἐναγώνιος ) is often used in opposition to ‘historical’ ( ἱστορικός ) and ‘narrative’ (διηγηματικός ), which imply a detachedness on the part of the audience: Long. Subl. 25, for instance, posits that the use of the historical present can turn a passage from a detached ‘narrative’ ( διήγησις ) into an ‘engaging event’ ( ἐναγώνιον πρᾶγμα ). Cf. also Long. Subl. 9.13, comparing the relaxedness of Hom. Od. to the action- packed drama of Hom. Il. , and Dion. Hal. Ant. Rom. 1.8.3, juxtaposing historical narrative for ‘undisturbed entertainment’ (ἀόχλητος διαγωγή ) and engaging oratory: see Ooms and De Jonge (2013) 102–104.

HERO VERSUS ZERO hidden in the middle of a long hyperbaton (between τῶν and ἐθνῶν), the preposition ‘between’ ( μεταξύ ) is postponed, and the word ‘northern’ ( βορέαν ) is separated from its governing noun ‘wind’ ( ἄνεμον ). In the final part of the sentence, the word ‘so-called’ (καλούμενον ) has been removed from its natural place: while it usually stands directly behind the name to which it refers, Dionysius has placed it the beginning of the word group, holding back the name itself until the very end ( ἐς καλούμενον πόντον Εὔξεινον ). 165 It is obvious that Dionysius does not attack Hegesias’ short, jerky phrases in this pastiche: after all, the phrases in Herodotus’ original and in the Thucydidean rewriting are of equal length. Yet, Dionysius may imitate the Magnesian’s preoccupation with a limited number of rhythmical cadences: five out of the seven clauses end on Hegesias’ favorite rhythms, that is, either a final ditrochee ( ― ᴗ | ― ―), appearing three times ( γένος δὲ Λυδός, and τύραννος ἐθνῶν, and Παφλαγόνων μεταξύ), or a cretic + spondee ( ― ᴗ ― | ― ―), appearing twice ( υἱὸς ἦν Κροῖσος and πόντον Εὔξεινον ).166 To compare, the Thucydidean version only exhibits such clausulae three times, and Herodotus’ original once.167 Like Cicero, then, Dionysius may have been aware of Hegesias’ best-loved rhythmical cadences. In his Hegesian recasting, the Greek critic imitates the Magnesian’s preparedness to arrange his words in extreme, nonsensical ways in order to create the artistic effects that he so desperately

165 Cf. Weil (1844) 129–130 on the first three clauses of Dionysius’ Hegesian rearrangement: ‘Le nom d’Alyattès, placé au commencement de la phrase et suivi de la particule μέν , se prononce avec un accent fort qui ne lui convient guère. (…) Au reste, Alyattès, fût-il un père très illustre, on ne fait pas ici un panégyrique, mais une généalogie. Si le nom d’Alyattès est trop mis en relief, celui de Crésus est trop effacé, et il ne pourrait guarder cette place qui s’il eût été question de Crésus dans les phrases précédentes, et que son nom ne fût répété que pour la clarté.’ De Jonge (2005) 478 (= De Jonge (2008) 388) submits that ‘this metathesis is associated with the “Asianic” style, to which the Atticist Dionysius strongly objects.’ 166 For Hegesias’ favorite rhythms and Cicero’s familiarity with them, see section 2.4.2 above. For an explanation of my approach to distinguishing phrases and scanning prose, see ibid. n. 120–122. The hunt for well-loved clausulae may explain the strange position of τύραννος in the second phrase, and the strange arrangement of καλούμενον πόντον Εὔξεινον in the final phrase. As far the two phrases that do not end either on a dicretic or on cretic + spondee, one has a final paean ( ᴗ ᴗ ᴗ ―, ἐξίησιν ἄνεμον ), which was Aristotle’s favorite (Rh. 3.8.6), and which De Groot (1921) 64 recognizes as one of Hegesias’ favorite clausulae. The other ends on a hypodochmiac ( ― ᴗ ― ᴗ ―), which is often attested in Greek imperial prose according to Hutchinson (2015) 789, but which I have not seen in the other fragments of Hegesias. 167 In Dionysius’ Thucydidean rearrangement, we encounter: υἱὸς μὲν Ἀλυάττου (― ᴗ ᴗ ᴗ | ― ―), γένος δὲ Λυδός (― ᴗ | ― ―), καλούμενον πόντον ( ― ᴗ ― | ― ―). In Herodotus’ Atticized original, we find only καλούμενον πόντον (― ᴗ ― | ― ―). The original Ionic-Greek text, which has Ἀλυαττέω , μεσημβρίης . βορῆν and καλεόμενον contains none of Hegesias’ favorite clausulae.

CHAPTER TWO craves.168 In sum, Dionysius not only takes exception to Hegesias’ ignoble, short-syllabled rhythms but also to his degenerate, logic-defying word order. In both instances, the Greek critic repeats the point that Cicero and Agatharchides had made before him about the Magnesian’s style: it has no place in a serious, dignified discourse.

2.4.4 Summary: the Principal Stylistic Vices of Hegesias What do the ancient discussions of Hegesias’ supposed stylistic ineptitude teach us about the authors’ views of style? As we have seen, Agatharchides, Cicero and Dionysius each underline different aspects of the Magnesian’s remarkable prose: Agatharchides primarily censures his figurative language, Cicero condemns his choppy, verse-like phrases, and Dionysius attacks his outrageously affected word order and his frequent clusters of short syllables. Despite these apparent differences of focus between the extant critics and rhetoricians, we may still recognize at least three shared points about the man from Magnesia:

1. Whereas Demosthenes’ oeuvre was considered a veritable catalogue of stylistic theory put into practice, Hegesias’ prose is not awarded a place in the stylistic categories of the surviving discussions of prose style. Neither Cicero nor Dionysius mention him in their three-style divisions: apparently, the inadequacy of the ‘Hegesian type of style’ (Hegesiae genus , Ἡγησιακὸν σχῆμα ) defies classification. In contrast to Demosthenes’ brilliant versatility, Hegesias allegedly knows only a handful of ostentatious tricks and he never loses an opportunity to showcase them, rendering his style formulaic and monotonous.

2. The critics and rhetoricians are unanimously shocked by the incongruity between Hegesias’ stylistic attributes and the serious topics that he addresses. Perpetually

168 The three brief samples of Hegesias’ prose (= F18–20 Jacoby) that Dion. Hal. Comp. 4.11 cites immediately after his pastiche also show a tendency to favor stylistic flourishes over a clear exposition of the substance. In F18, the arrangement creates alliteration, assonance, rhyme, and the clausula paean + spondee ( ― ᴗ ᴗ ᴗ | ― ―): ‘After a good festival another good one we celebrate’ ( ἐξ ἀγαθῆς ἑορτῆς ἀγαθὴν ἄγομεν ἄλλην ). In F19, Hegesias introduces himself in a strangely ordered sentence that resembles Dionysius’ Hegesian version of the introduction of Croesus: ‘From Magnesia am I, the mighty land, a Sipylean’ ( ἀπὸ Μαγνησίας εἰμὶ τῆς μεγάλης Σιπυλεύς ). F20, lastly, exhibits a highly obscure syntax, ending its second colon on a ditrochee ( ― ᴗ | ― ―), and its third on cretic + spondee ( ― ᴗ ― | ― ―): ‘It was no small drop that into Theban waters Dionysus spew: for sweet it is indeed, but it makes men mad’ ( οὐ γὰρ μικρὰν εἰς Θηβαίων ὕδωρ ἔπτυσεν ὁ Διόνυσος· | ἡδὺ μὲν γάρ ἐστι , | ποιεῖ δὲ μαίνεσθαι ).

HERO VERSUS ZERO

hunting for frivolous rhythms and light-hearted balance, he was considered unable to handle the lofty, life-and-death issues of practical oratory, where the Attic orator Demosthenes was thought to reign supreme. The abuse that he receives in our sources reflects his reputation as an unserious author: he is described as juvenile, effeminate and insane, arousing laughter when pathos is required.

3. According to Cicero and Dionysius, Hegesias’ prose betrays an utter lack of talent for word arrangement. Intriguingly, both scholars provide apposite burlesques of their antimodel’s composition. Cicero, for one, mockingly imitates the ‘verselets’ (versiculi ) that can indeed be found in Hegesias’ extant fragments. Dionysius, in addition to his criticism of the Magnesian’s ignoble rhythms, offers a striking parody of his affected word order. To be brief, word arrangement was not only considered a pre-eminent source of Demosthenes’ success but also of Hegesias’ failure.

In the surviving Greek and Latin discussions of prose style, then, Hegesias is presented as a virtual anti-Demosthenes: the Magnesian author is thought to have failed on the very points on which the Athenian hero is said to have triumphed. Hence, the Lydian orator and historian was a perfect prototype of ‘Asian’ style ( Ἀσιανός , Asianus, Asiaticus ), which, as we will see, was set up as the polar opposite of everything that the critics and rhetoricians of Late- Republican and Augustan Rome liked about Attic oratory (sections 5.2 and 5.3). The harsh ostracism of Hegesias offers us a rare concrete glimpse into the shared Greek and Roman conceptions of the ‘bad’ style of Hellenistic Asia, highlighting, by contrast, what Cicero, Dionysius and their colleagues love about the ‘good’ style of classical Athens.

2.5 Conclusion In this chapter, we have seen that Cicero and Dionysius both select Demosthenes as the ultimate champion of brilliant prose, while they both present Hegesias of Magnesia as the prime exponent of bad taste. The prototypicality of Demosthenes and Hegesias is connected to the dominant prejudice about their places of birth and the eras in which they were active: Demosthenes lived and worked in Athens at the end of the Classical Era, whereas Hegesias was born in Asia, flourishing at the beginning of the Hellenistic Era. I have argued, however, that the lavish praise of Demosthenes and the fierce criticism of Hegesias can not only be explained as products of classicizing cultural bias, but also as the results of serious investigations into the properties of their styles: we have seen that Cicero and Dionysius stress

CHAPTER TWO the same virtues in the Athenian’s oratory, likewise emphasizing the same vices in the Magnesian’s prose. What is more, the former’s strengths are argued to be the latter’s weaknesses. Cicero and Dionysius not only reveal to us what they consider the distinctive qualities of good and bad prose, but, since several important elements from their discussions can be found in other contemporary sources as well, they also shed a telltale light on the prevalent aesthetic tastes in Late-Republican and Augustan Rome. The following are Cicero’s and Dionysius’ three crucial criteria for distinguishing between admirable and deplorable prose: style should (1) be adaptable to suit any circumstances, (2) hold up particularly in the treatment of the serious, dignified issues of real- life oratory, and (3) pay special attention to the art of word arrangement. These shared views are integral features of the common critical framework on which the Greek and Roman discourse on prose style is built. We should note, however, that each of the critics and rhetoricians could bend these basic parameters to suit their individual aesthetic programs. Concerning rhythm, for instance, Cicero contrasts Demosthenes’ forceful thunderbolts to Hegesias’ minced verse-like fragments, whereas Dionysius plays off the nobility of the former’s long-syllabled rhythms against the degeneracy of the latter’s short-syllabled feet. In the remainder of this dissertation, I will make a closer study of the flexible application of the shared stylistic repertoire that we encountered in the present chapter. The next chapter, to begin with, will focus on the various threefold divisions of style that the Greek and Roman authors use to classify classical literature and to analyze Demosthenes’ flexible prose. Sadly, the sorry tale of Hegesias ends here: as he is denied any official status in the extant stylistic models, the man from Magnesia will only reappear occasionally in the margins of my subsequent analyses.

Chapter 3 THE PLAIN , THE GRAND AND THE IN-BETWEEN : THE DOCTRINE OF THREE STYLES AS A VERSATILE CRITICAL TOOL

3.1 Introduction In the previous chapter, we have learnt that Cicero and Dionysius both consider versatility a prime feature of excellent literature: they attribute this cardinal virtue to Attic prose in general and to Demosthenes’ oratory in particular. In addition, we have seen that they adopt similar, though far from identical, threefold stylistic divisions to make sense of the resourceful flexibility of their literary heroes, applying it not only to Demosthenes, but also to other prose authors such as Thucydides, Lysias and Isocrates, and, in Dionysius’ case, to various poets, from Homer to Hesiod, from Pindar to Sappho, and from Aeschylus to Euripides. In this chapter, we will take a closer look at these three-style classifications, exploring the relevant similarities and differences between the extant versions of the triple scheme, paying special attention to the cross-connections between the surviving Greek and Latin tripartitions. The doctrine which holds that there are ‘three types of style’ (χαρακτῆρες τῆς λέξεως , genera dicendi )—one plain, one intermediate and one grand—is attested for the first time in the first century BC. The earliest extant articulations of the theory are in Latin, in the fourth book of the anonymous Rhetorica ad Herennium and in Cicero’s rhetorical treatises, especially in his Orator. Varro also deals with the theory, but the evidence is too vague to make confident statements about his teachings on the topic. 1 It is usually assumed that these Latin sources reflect an older Hellenistic tradition, whose precise origins are now lost in the meager record of that era. 2 In the extant sources on Greek stylistic theory, however, we typically do not find threefold stylistic divisions, but rather fourfold categorizations (Demetrius, Philodemus) and more elaborate systems (Hermogenes, pseudo-Aristides). 3 Our

1 Rhet. Her. 4.11–16, Cic. De or. 3.177, 3.199, 3.212, Orat. 20–139, Opt. gen. 1–6. According to Gell. NA 6.14.6 (= Varro fr. 80 Wilmanns), Varro links Latin dramatic poets to all three styles: he connects the tragic poet Pacuvius to ‘abundance’ (ubertas ), the satirist Lucilius to ‘slenderness’ ( gracilitas ), and the comic poet Terentius to the ‘middle style’ ( mediocritas ). In the passage, it is not entirely clear at which point Gellius stops paraphrasing Varro: section 3.3 n. 51 below. Varro’s lost work On Types (Περὶ χαρακτήρων ), ascribed to him by Charisius Gramm. 2.246 Barwick, dealt with the grammatical formation of words, not with types of style. 2 Cf. section 1.2 above on source criticism and the hellenocentric approach to ancient stylistic theory. 3 Demetr. Eloc. recognizes the ‘plain’ ( ἰσχνός ), ‘smooth’ ( γλαφυρός ), ‘grand’ ( μεγαλοπρεπής ) and ‘forceful’ (δεινός ) registers. The relationship between Demetrius’ fourfold scheme and the extant three-style systems is the

CHAPTER THREE first Greek source for the theory of three styles is Dionysius of Halicarnassus, specifically his treatises On Demosthenes and On the Arrangement of Words . After Dionysius, the three-style formula is rarely attested in Greek sources. 4 Thus, threefold divisions of style seem to have been a typical feature of the stylistic discourse in Late-Republican and Augustan Rome. Later Latin critics, such as Quintilian and Augustine, retain the division, mainly on the authority of Cicero: this tradition continues until the end of the Middle Ages. 5 Modern scholars are interested in the ancient theory of three styles for roughly two reasons: one concerns the origin of the theory, the other its usefulness as a critical tool. First, there is a long-standing and still undecided debate about the ultimate provenance and early development of the theory. 6 The dichotomy between the grand and the plain, which forms the subject of some discussion: Kennedy (1963) 284, Calcante (2000) 142–152, id. (2004) 100–103 and Marini (2007) 179–183 claim that Demetrius’ division is an adaptation of an older threefold system, but others, such as Schenkeveld (1964) 54–55 and, abandoning his earlier view, Kennedy (1989) 196, have more convincingly argued that Demetrius does not refer to any three-style system. Rather, as Demetr. Eloc. 36 explicitly states, the four-style doctrine is an elaboration of the standard two-style view that is dominant in Greek criticism before Rome and that also forms the basis of the surviving threefold doctrines: see the next paragraph below. Philod. Rhet. 4 col. 4.2–5 p. 165 Sudhaus seems to distinguish between four ‘forms’ ( πλάσματα ), viz., ‘plainness’ (ἰσχνότης ), ‘smoothness’ ( γλαφυρότης ), ‘grand writing’ ( ἁδρογραφία ) and a fourth type whose name is illegible apart from an initial μ: Sudhaus conjectures the word μεσότης , which he draws from the three-style formula, whereas Radermacher (1899a) 361 proposes μέγεθος , which connects Philodemus’ division with Demetrius’ scheme, provided that Philodemus’ ἁδρογραφία corresponds to Demetrius’ χαρακτὴρ δεινός . On fourfold constructs in ancient thought, see Usener (1913). In the second century AD, Hermog. Id. and [Aristid.] articulate complex systems of many ‘types’ ( ἰδέαι ) of style: see Wooten (1987) and Rutherford (1998). 4 See Dion. Hal. Dem. 1–34 for the three ‘types of diction’ ( χαρακτῆρες τῆς λέξεως ), and ibid. 35–52 and Comp. 21–24 for the three ‘types of arrangement’ ( χαρακτῆρες τῆς συνθέσεως ): cf. sections 3.2.1 and 3.3 below for the interconnections between both systems. [Plut.] Vit. Hom. 172–173 and Procl. Chr. (in Phot. Bibl. cod. 239) both articulate three-style divisions. Nünlist (2009) 219–221 notes that the three styles have had little influence on the scholia: a possible reference can be found in schol. bT Il. 21.257–62a which states that Homer ‘makes a transition from the grand style to the plain and florid’ ( ἀπὸ τοῦ ἁδροῦ ἐπὶ τὸ ἰσχνὸν ἔρχεται καὶ ἀνθηρόν ). Cf. section 3.2.2 n. 46 below on Quint. Inst. orat. 12.10.58, who describes the middle style as ‘florid’. See also section 3.3 below on Odysseus, Menelaus and Nestor as stylistic prototypes. Rutherford (1998) 10–18 discusses the connections between Dionysius’ three styles and the intricate stylistic divisions in Hermog. and [Aristid.]. 5 Quint. Inst. orat. 12.10.58–72, August. Doctr. christ. 4.32–58. For Augustine’s treatment of the three styles, see Sluiter (1999) 256–260. For the late-antique and medieval tradition, see Quadlbauer (1962): cf. n. 6 below. 6 The modern discussions of the ancient history of the three styles adopts various approaches. Schmid (1894), for one, reconstructs the tradition, working his way back from Procl. Chr. and arguing that the theory originates in the Stoa. Hendrickson (1905), conversely, explores the Peripatetic background of the theory: cf. esp. section 3.5 below. Next, Wehrli (1946) addresses the connection between poetic and rhetorical theory, tracing the doctrine

THE PLAIN , THE GRAND AND THE IN-BETWEEN backbone of the three-style doctrine, deeply influences Greek literary criticism from Homer onwards: the poet already portrays Menelaus as speaking plainly, whereas he ascribes to Odysseus a more elevated style. A similar contrast is drawn up in the famous contest between Euripides and Aeschylus in Aristophanes’ Frogs. 7 The concept of an intermediate style seems to have been a later addition to this binary scheme. The Peripatetic school has been extremely influential with its focus on the appropriate mean (μεσότης ) between the two stylistic extremes: in the third book of his Rhetoric , Aristotle prescribes that style should be ‘neither base nor elevated above the subject matter, but appropriate’ ( πρέπουσα ). Indeed, the criterion of ‘appropriateness’ ( τὸ πρέπον , decorum, aptum ) is a major ingredient in the extant three- style divisions of Late-Republican and Augustan Rome: the surviving critics and rhetoricians insist that style should be appropriate to the purpose of the speech, to the subject matter under discussion, to the particular section of the speech, to the moral character of the speaker, and to the composition of the audience. 8 The hottest issue in the secondary literature about the origin of the three styles is the debate as to whether Aristotle’s successor in the Lyceum, Theophrastus, was the first to develop a full-blown theory of three styles. 9 The matter hinges almost entirely on the

back to the debate about φύσις and τέχνη in the fifth century BC. Augustyniak (1957) addresses the interplay between theories of three styles and fourfold doctrines. Quadlbauer (1958) and (1962) traces the history of the threefold division from Homer until the Middle Ages: his approach, however, is rather narrow, mostly limited to issues of grammar and sentence structure. For the supposed Theophrastean origin, see the next paragraph below. 7 See section 3.3 below on Hom. Il. 3.212–224 (Menelaus and Odysseus) and ibid. 1.247–249 (Nestor). O’Sullivan (1992) and Hunter (2009) 29–36 establish the strong connections between the agon in Ar. Ran. 830– 1117 and various later sources of criticism, such as Arist. Rh. , Demetr. Eloc. , Dion. Hal. and Long. Subl. Note that the dichotomy between simplicity and elevation is also present in Demetrius’ fourfold scheme: at Eloc. 36, he explains that the ‘plain’ ( ἰσχνός ) and ‘grand’ ( μεγαλοπρεπής ) styles ‘stand, as it were in polar opposition and contrast’ ( ὥσπερ ἀνθέστατον καὶ ἀντίκειθον ἐναντιωτάτω ). 8 Arist. Rh. 3.2: ‘Let the virtue of style be defined as ‘to be clear’ (…) and neither flat nor above the dignity of the subject, but appropriate’ ( ὡρίσθω λέξεως ἀρετὴ σαφῆ εἶναι ... καὶ μήτε ταπεινὴν μήτε ὑπὲρ τὸ ἀξίωμα ἀλλὰ πρέπουσαν ). Transl. Kennedy (1991). See section 3.5 below on the Peripatetic notion of the stylistic mean in Dionysius’ three-style doctrines. Our sources present the middle style as a later addition to the existing two opposite styles: cf. Cic. Orat. 21 ( interiectus inter hos medius ), Dion. Hal. Dem. 3.1 ( μικτή τε καὶ σύνθετος ἐκ τούτων τῶν δυεῖν), Quint. Inst. orat. 12.10.58 ( tertium addiderunt ). The virtue of appropriateness is stressed by Cic. De or. 3.199, Orat. 70–74; Dion. Hal. Dem. 34.5, Comp. 12.10, Pomp. 3.20; Quint. Inst. orat. 12.10.69–72. 9 Before Hendrickson (1904) and (1905), it was generally believed that Theophrastus had been the founding father of the theory of three styles: see e.g. Radermacher (1899b). After Hendrickson’s seminal articles, however, there is an ongoing debate about the Theophrastean origin of the doctrine: Kroll (1907), Mayer (1910),

CHAPTER THREE interpretation of one particular passage in Dionysius’ On Demosthenes , which discusses the origin of the middle style: Dionysius credits Theophrastus with the opinion that Thrasymachus of Chalcedon was the inventor of the intermediate register. It is on the basis of this testimony that many scholars have argued that Theophrastus must have discussed the three styles in his lost work On Style (Περὶ λέξεως ). Yet, it is equally possible that the Peripatetic scholar recognized in Thrasymachus’ style a ‘source for the mean’ ( πηγή τις τῆς μεσότητος ), as Dionysius puts it, an appropriate mean between excess and deficiency, without acknowledging the existence of three separate stylistic levels. 10 Moreover, it is hard to imagine that the philosopher would have approved of a style that is anything else than an appropriate mean, or, to put it in John Hendrickson’s words: ‘The μεσότης was to Theophrastus not a style, but the style.’ 11 There is very little material except for the passage from Dionysius to support the thesis that Theophrastus discussed a three-style theory, nor indeed is there any hard proof for its existence prior to the first century BC. 12

Augstyniak (1957), Kennedy (1956), (1957b), (1963) 278–282, though he later changed his position, Quadlbauer (1958) 64–71 and Janko (2000) 156–157 maintain that the Peripatetic was the first to put forward the theory, while Stroux (1912) 88–104, Grube (1952), (1965) 107–108, Innes (1985) 260–263, Kennedy (1989) 195 and Fortenbaugh (2005) 273–278 argue that the philosopher knew no threefold theory. Cf. the notes below. 10 Dion. Hal. Dem. 3.1 (= Theophr. fr. 685 Fortenbaugh): ‘The third kind of style was a mixture formed by combining the other two; whether the person who joined them together and brought it to its present splendor was Theophrastus of Chalcedon, as Theophrastus thinks, or somebody else, I cannot say’ ( τρίτη λέξεως ἰδέα ἦν ἡ μικτή τε καὶ σύνθετις ἐκ τούτων τῶν δυεῖν, ἣν ὁ μὲν πρῶτος ἁρμοσάμενος καὶ καταστήσας εἰς τὸν νῦν ὑπάρχοντα κόσμον εἴτε Θρασύμαχος ὁ Χαλκηδόνιος ἦν, ὡς οἴεται Θεόφραστος , εἴτε ἄλλος τις , οὐκ ἔχω λέγειν ). The view that Theophrastus advocated not three styles but a stylistic mean only is first put forward by Hendrickson (1904) and subsequently adopted by many other scholars (cf. n. 9 above): important contributors to this thesis are Grube (1965), who shows that Dionysius’ λέξις refers specifically to diction (cf. section 3.2 below) and that we, hence, cannot attribute Theophrastus a conception of middle style but rather the notion of ‘a mean or middle diction, that is, the development of a prose vocabulary’, and Innes (1985) 260–263, who adds that the philosopher advocated the mean not only in diction, but also in rhythm and probably sentence structure. See Fortenbaugh (2005) 273–278 for an overview of the secondary literature on Dion. Hal. Dem. 3.1–5: Fortenbaugh considers the possibility that the phrase ‘a source for the mean’ ( πηγή τις τῆς μεσότητος ) was used by Theophrastus himself in relation to the style of Thrasymachus. On the connections between the Peripatetic mean and Dionysius’ three styles, see section 3.5 below. 11 Hendrickson (1904) 140. 12 Mayer (1910) 1–50 finds additional evidence for the Theophrastean origin of the three-style doctrine in Dion. Hal. Isoc. 3.1, reporting that the Peripatetic lists three means to achieve ‘grandeur, dignity and impressiveness’ (τὸ μέγα καὶ σεμνὸν καὶ περιττόν ). As Grube (1952) 266 argues, however, the quoted adjectives are used in a general sense of distinction or excellence in a writer: they do not indicate a familiarity with the notion of a grand

100

THE PLAIN , THE GRAND AND THE IN-BETWEEN

An unfortunate side effect of the scholarly focus on the prehistory of the theory of three styles is that the critics and rhetoricians in Rome are often imagined as rather passive recipients of older doctrines. In the search for the primordial inventor of the theory, the originality and variety of the extant sources tend to be overlooked. 13 The only aspect of the theory that is generally considered to be an innovation of the first century BC is the connection that Cicero makes between the three types of style and the three so-called ‘functions of the orator’ ( officia oratoris ), that is, ‘teaching, placating and moving’ ( docere, conciliare, movere ), or, alternatively, ‘proving, entertaining and swaying’ ( probare, delectare, flectere ). This link becomes a standard feature in Latin rhetorical theory, but it is not found in Dionysius. 14 As I explained in the introduction to this dissertation (section 1.2), I will adopt a synchronic perspective to the stylistic theories in the works of Cicero, Dionysius and their contemporaries, focusing not on the origins or afterlife of their doctrines, but rather on the underlying goals and motivations. No matter when the theory of the three styles was first articulated, I will focus on its various applications in the Greek and Latin stylistic discourse of Late-Republican and Augustan Rome, when the threefold doctrine was one of the go-to critical tools for the analysis of literary prose. A second line of inquiry in modern studies of the three styles focuses on the usefulness of the theory: are the threefold systems at all useful for the stylistic analysis of classical Greek and Latin prose? Can Cicero’s articulation of the doctrine, for instance, be applied to his own speeches? Some scholars have answered these questions in the affirmative, 15 but the style, cf. Nünlist (2009) 221. Kennedy (1957b) cogently refutes various assertions by others, who claim to have found traces of the theory of three styles in Greek sources from the Classical and Hellenistic eras. 13 An extreme example of this attitude can be found in Peterson (1891) 44 on Quint. Inst. orat. 12.10.58: ‘This threefold division of style, ascribed to Theophrastus, was generally recognized in Greece after the latter part of the fourth century B.C. (…) It is adopted in Cornif. ad Herenn. (…) and is carefully explained by Cicero in the Orator (…) Dion. Hal. (probably following Theophrastus περὶ λέξεως ) has the same division (…). Quintilian repeats this.’ The italics are mine. 14 The first attestation of the fusion of the three styles and the three functions of the orator is Cic. Orat. 69: ‘There are as many types of style as there are functions of the orator’ ( Quot officia oratoris tot sunt genera dicendi ). According to Douglas (1957), this is a ‘Ciceronian contribution to rhetorical theory’. See section 3.2, esp. table 3 below. For later articulations of the theory, see Austin (1948) 199–200 on Quint. Inst. orat. 12.10.59, and Sluiter (1999) 256–260 on August. Doctr. christ. 4.34. 15 For an optimistic view about the usefulness of the three-style doctrine for the analysis of Cicero’s speeches, see Adamik (1995), who shares the positive appreciation of the French scholars Laurand (1907) and Marouzeau (1935): ‘Theoretical and practical informations of the Latin rhetoricians are useful for analyzing their speeches, because they help to recover the intentions of their authors.’

101

CHAPTER THREE predominant scholarly opinion is that the theory is of limited use in the study of ancient literature. Dionysius and Quintilian, two major exponents of the doctrine, already admit that the threefold division is inherently inaccurate in the face of the sheer myriad of potentially and actually existing styles. 16 In modern literature, the theory of three styles is often considered a cliché and not very helpful at all: it has been called ‘rather meaningless’ (Grube), ‘at best a widely accepted pedagogic convenience’ (Russell), ‘virtually useless as a tool for serious literary criticism’ (Douglas), ‘not very illuminating and even constrictive’ (Wooten), and, in milder terms, ‘more or less true to the broad facts; but it falls short in precision when it comes to details’ (Powell). 17 Indeed, vague and broad concepts such as ‘plain’, ‘intermediate’ and ‘grand’ seem out of place in a nuanced and refined discussion of style. In a word, the three-style formula is seen by many scholars as both unoriginal and unpractical. Concerning the presumed unoriginality, I will contend that the opposite is true. I will show that the three authors who deal with the topic most extensively (the author of Rhetorica ad Herennium, Cicero, Dionysius) do not simply copy a rigid, monolithic doctrine of three styles from their illustrious predecessors: on the contrary, the theory proves to be highly flexible and the categories involved turn out to be exceedingly fluid (section 3.2). As, during the first century BC, the critics and rhetoricians in Rome became increasingly obsessed with the literature and style of Classical Greece, this malleable theory came to be used as a convenient framework to analyze and categorize the Greek literary heritage (section 3.3). The surviving three-style systems vary considerably: Dionysius, for instance, argues, as we have seen, that the middle style is the best style and that its ultimate champion is Demosthenes (section 2.3.3), whereas Cicero considers the middle style the least important of the three registers, proposing the ambivalent Peripatetic philosopher Demetrius of Phalerum as its prototypical exponent. Cicero’s grand style, in addition, is represented by the orator Pericles, while Dionysius selects the historian Thucydides as the exemplary author of the elevated style. 18

16 Quint. Inst. orat. 12.10.66: ‘Eloquence is not limited to these three patterns’ ( sed neque his tribus quasi formis inclusa eloquentia est ). Dion. Hal. Comp. 21.1: ‘I hold the view that there are very many distinct forms of word arrangement, which can be included neither in a comprehensive view nor in a detailed reckoning’ ( ἐγὼ τῆς συνθέσεως εἰδικὰς μὲν διαφορὰς πολλὰς σφόδρα εἶναι τίθεμαι καὶ οὔτ’ εἰς σύνοψιν ἐλθεῖν δυναμένας οὔτ’ εἰς λογισμὸν ἀκριβῆ). 17 Grube (1952) 267, Russell (1964) xxxv, Douglas (1973) 115, Wooten (1989) 587, Powell (2013a) 53. 18 For Cicero’s and Dionysius’ categorizations of Greek authors according to the three styles, see section 3.3 table 7 below.

102

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Next, I will address the issue of the practical application of the theory. Although I must admit that the three styles seem of limited use for the analysis of the artistic subtleties inherent in Greek and Latin literary prose, we will see that the theoretical exposés of Cicero and Dionysius do serve a practical purpose, albeit a different one: they use the triple scheme not in every part of their oeuvre, adducing it only when they need it to prove a particular point. Cicero, for one, barely mentions the three styles in his De Oratore , while in his Brutus he propagates a system of two rather than three styles; yet, in his Orator , he relies heavily on a threefold system, which he skillfully puts to use in his polemic against the self-proclaimed Attic orators in Rome (section 3.4). Similarly, Dionysius does not refer to the theory in his works on Lysias, Isocrates, Isaeus or Thucydides, but he dwells on it at length in his On Demosthenes , where he uses it to demonstrate the stylistic versatility and universal appeal of Demosthenes, and again in On the Arrangement of Words , where he makes a similar point about stylistic adaptability (section 3.5). Hence, the differences between the threefold systems that Cicero and Dionysius expound can to a large extent by attributed to the different programs that the authors articulate in their works. There is, lastly, a more general point to make about the practical use of the three-style doctrine (section 3.6). Both Cicero and Dionysius claim, as we have already seen, that the orator who is proficient in all three styles will be successful in the real-life oratory of forensic and deliberative speeches (sections 2.3.2 and 2.3.3). Thus, Cicero and Dionysius do not only use the theory of three styles to analyze the shared literary heritage of classical Greece, but they also use it to point the way to rhetorical success in the contemporary Roman world. In a practical vein, both the Roman rhetorician and the Greek critic take the makeup of the audience into account in their discussions of the three styles: the audience that Cicero evokes, however, is not identical to the public that Dionysius refers to. The former connects his three- style theory specifically to the Roman masses on the Forum, while the latter generally makes reference to congregations of Greek crowds in the ‘public assemblies’ ( ἐκκλησίαι ) and ‘law- courts’ ( δικαστήρια ).19 We will see that a Greek audience, apparently, calls for a different application of the three styles than a Roman audience.

19 I do not argue that Dionysius’ works are not meant for Roman readers: Comp. was addressed to the young Roman aristocrat Metilius Rufus, and Thuc. to the Roman historian and lawyer Q. Aelius Tubero. For Dionysius’ audience, see section 1.5 above. It is, however, a well-established fact that Dionysius’ literary essays only refer to Greek literature and, as we will see in section 3.6 below, to the Greek contexts of that literature. Metilius Rufus became proconsul of Achaea and, perhaps, legate of Galatia, where he had many opportunities to put Dionysius’ teachings about rhetoric into practice: cf. section 1.5 n. 91 above.

103

CHAPTER THREE

3.2 Versatile Systems of Three Styles The versatility of the threefold formula is reflected in the vocabulary of its users. Quintilian provides the locus classicus for the Greek and Latin terminology of the three styles: ‘One style ( genus ) is defined as subtile (Greek ἰσχνόν ), a second as grande and robustum (Greek ἁδρόν ), and to these has been added a third, called by some medium ex duobus , and by other floridum (for the Greeks call it ἀνθηρόν ).’ 20 The Greek antithetical pair ἰσχνόν versus ἁδρόν and the Latin translation subtile versus grande refer to opposite appearances of the human body: one style is imagined as a ‘thin’ or ‘slim’ body, while its counterpart is described as ‘thick’ or ‘stout’. 21 The middle style is usually regarded simply as ‘intermediate’ between these two extremes, although it is sometimes thought of as ‘flowery’.22

Varro Cic. De or. 3.177, 3.199, 3.212, Rhet. Her. 4.11–16 (Gell. NA 6.14.5) Opt. gen. 1–6, Orat. 20–139 orator tenuis , subtilis (thin) , acutus gracilitas figura adtenuata 1 (astute) , submissus (calm) , humilis (low) , (slender) (thin) brevis (concise) orator mediocris , medius , interiectus mediocritas figura mediocris 2 (intermediate) , temperatus (moderate) , (intermediate) (intermediate) modicus (modest) orator gravis (heavy) , plenus (full) , ubertas figura gravis 3 copiosus (rich) , vehemens (passionate) , (abundant) (heavy) grandiloquus (grand)

Table 1: Latin terminology for the three styles (first century BC)

20 Quint. Inst. orat. 12.10.58: Namque unum subtile, quod ἰσχνόν vocant, alteram grande et robustum, quod ἁδρόν dicunt, constituunt, tertium alii medium ex duobus, alii floridum (namque id ἀνθηρόν appellant ) addiderunt . Russell (1964) xxxvi and Lausberg (2008) 472–475 offer concise overviews of the ancient terminology of the three styles, relying heavily on Quintilian’s terminology: convenient as they are, these synopses reinforce a positivist approach to the diverse source material (cf. section 1.2 n. 13 above) giving the false impression that there was a unified system that authors adopted wholesale. 21 See Sluiter (2010) 30–36 for the ancient conception of language and literature as resembling bodies or organisms. Iodice di Martino (1986) 26–27 offers a list of passages from Cicero’s rhetorical works in which the author applies the metaphor of the human body to oratory. Dugan (2005) 270–279 shows that ‘the body is the underlying metaphor that governs the Orator ’s discourse on style’. 22 For the flowery style, cf. Cicero’s conception of the middle style: see section 3.2.2, esp. n. 46 below.

104

THE PLAIN , THE GRAND AND THE IN-BETWEEN

‘Types of diction’ ‘Types of arrangement’ (χαρακτῆρες τῆς λέξεως ) (χαρακτῆρες τῆς συνθέσεως ) Dion. Hal. Dem. 1–34 Dion. Hal. Comp. 21–24, Dem. 35–52 λέξις λιτή , ἀφελής (smooth, plain), 1 σύνθεσις γλαφυρά (smooth, polished) ἰσχνή (thin) λέξις μικτή (mixed), σύνθετος σύνθεσις εὔκρατος (well-blended), κοινή 2 (compounded), μέση , μεταξύ (common), μέση (intermediate) (intermediate) λέξις ὑψηλή (elevated), περιττή 3 σύνθεσις αὐστηρά (rough) (extraordinary), ἐξηλλαγμένη (unusual)

Table 2: Dionysius’ terminology for the three styles

Quintilian’s neat terminology may give the impression that there existed a uniform Greek and Latin three-style system. Yet, as the foregoing two tables illustrate, the sources from the first century BC display a rather diverse nomenclature: apparently, ancient critics and rhetoricians did not simply subscribe to a standard doctrine, but they could adopt and adapt the three-style formula according to their own preferences. Varro, the author of Rhetorica ad Herennium , Cicero and Dionysius do not only apply various adjectives to each of the three styles, but their adjectives are also governed by different nouns. Rhetorica ad Herennium , for instance, distinguishes between ‘types’ ( figurae ) and sometimes ‘kinds’ ( genera ). In his Orator , Cicero applies the attributes to the orators themselves, although he also refers to ‘types’ (figurae ) and ‘kinds’ (genera ) elsewhere.23 Dionysius, lastly, has not one, but two systems of stylistic ‘types’ ( χαρακτῆρες ), one of which covers word selection, or diction (λέξις ), while the other is concerned with word arrangement, or composition (σύνθεσις ). The types of arrangement are also called the three ‘harmonies’ ( ἁρμονίαι ).24 The Latin three-style systems, conversely, comprehensively address all parts of style—selection, arrangement and figures of speech. 25

23 Rhet. Her. 4.11: ‘There are, then, three kinds of style, which we call types’ ( sunt igitur tria genera, quae genera nos figuras appellamus ). Note that the term figurae for ‘figures of speech’ (Greek σχήματα ) is used for the first time in Quint. Inst. orat : Cicero refers to figures as lumina (e.g. De or. 3.205) and Rhet. Her. uses exornationes (e.g. 4.18). As for Cicero’s terminology for the three styles, De or. 3.177 has genus , while he uses figura at ibid. 3.199 and 3.212. In Orat ., however, Cicero rather distinguishes between three types of orator: see e.g. Orat. 69, 100. On Gellius’ account of Varro’s threefold division, see section 3.1 n. 1, and section 3.3 n. 51. 24 Artés Hernández (2013) thinks that the terms λέξις and λόγος are often interchangeable in Dionysius’ works, both referring to style in general: he argues that Dionysius represents an intermediate stage between Philodemus

105

CHAPTER THREE

3.2.1 Malleable Dichotomies: the Stylistic Extremes All extant articulations of the three-style formula are built on the notion of two stylistic extremes—one plain and unadorned, the other grand and ornate. The third style is usually imagined as intermediate between these stylistic extremities. The author of Rhetorica ad Herennium , for one, calls the two opposites ‘thin’ ( adtenuata ) and ‘heavy’ ( gravis ), adding that both of these forms can degenerate into ‘neighboring faulty styles’ ( finitima et propinqua vitia ): the thin style is akin to the vicious ‘meagre’ type (exilis ), while the heavy style can easily become ‘swollen’ ( sufflata ). Incidentally, as we will see (section 3.2.2), the ‘intermediate’ style ( mediocris ) is akin to the so-called ‘slack’ style (fluctuans ). Interestingly, the author supplies self-written samples for the three virtuous types and their kindred vicious types, a clear demonstration of the practical applicability of ancient stylistic theory: for this reason, the section in Rhetorica ad Herennium about the three styles has been called a ‘textbook in Latin prose composition’. 26 Setting aside the discussion in the anonymous treatise, the present subsection will take a closer look at the formulations of the stylistic extremes in the works of Cicero and Dionysius. Like the anonymous author, Cicero uses the antithetical pair ‘thin’ ( tenuis ) and ‘heavy’ ( gravis ). Unlike the author of Rhetorica ad Herennium , however, Cicero adds moral descriptions to each of the stylistic types: as we have seen (table 1), he considers the orator of

(who still favors λέξις ) and Hermogenes and Longinus (who rather use λόγος ). Yet, as we will see in sections 3.2.1 and 3.2.2 below, Dionysius’ λέξις often (e.g. Lys. 8.2), though not always (e.g. Comp. 3.1), refers specifically to word selection. On Dionysius’ terminology for the types of arrangement, see Pohl (1968) 1–11: the terms ‘arrangement’ ( σύνθεσις ) and ‘harmony’ ( ἁρμονία ) not only evoke architectural imagery ( ἁρμόζω means ‘to join’), but also the acoustic, musical aspect of word arrangement. Cf. section 4.1 n. 7 below for earlier articulations of the theory of three types of word arrangement (Demetr. Eloc. , Philod. Poem. ). 25 See e.g. Rhet. Her. 4.11: the author notes that the grand style has ‘impressive words’ ( gravia verba ) as well as a ‘smooth and ornate arrangement’ ( levis et ornata constructio ). At ibid. 4.16, he adds that each type of style is embellished by appropriate figures of speech and thought. Likewise, Cic. Orat. 80 distinguishes between ‘embellishment in single words’ ( ornatus verborum simplicium ) and ‘embellishment in word combination’ (ornatus verborum collocatorum ), which comprises both rhythmical word arrangement and figures of speech. 26 Deneire (2004) provides a systematic discussion of the stylistic samples in Rhet. Her. , in each case reviewing word choice, rhythm, periodicity and other features. Deneire builds on the analyses of the same passages by Marouzeau (1921), Leeman (1963) 29–32 and Caplan (1969) 290–298. The notion of faulty styles can also be found in Demetr. Eloc. , who connects the ‘frigid style’ ( τὸ ψυχρόν ) to the grand type, and the ‘arid style’ (χαρακτὴρ ξηρός ) to the simple style. Cf. Gell. NA 6.14.5, who associates ‘pompous and bombastic speakers’ (sufflati atque tumidi ) with grandeur, ‘filthy and barren speakers’ ( squalentes et ieiuni ) with simplicity, and ‘unclear and ambiguous speakers’ ( incerti et ambigui ) with the middle style: cf. section 3.3 n. 51 below.

106

THE PLAIN , THE GRAND AND THE IN-BETWEEN the thin style ‘calm’ ( submissus ) and ‘astute’ ( callidus ), whereas he regards the orator of the heavy style as ‘passionate’ ( vehemens ), and the orator of the intermediate style ‘modest’ (modicus ) and ‘temperate’ ( temperatus ). Cicero’s diverse terminology reflects the variety of applications that he assigns to the three styles: according to the rhetorician, each of the three registers can be connected to subject matter, to various discourse functions and to the specific part of the speech. Thus, Cicero fits the theory of three styles neatly into other important parts of his rhetorical system. In each case, the guiding principle is ‘appropriateness’ ( τὸ πρέπον , decorum ). 27

Plain style Intermediate style Grand style Subject matter humilia, parva mediocria, modica alta, magna Discourse function probare delectare flectere Part of the speech narratio, argumentatio exordium peroratio

Table 3: The three styles within Cicero’s rhetorical system

First of all, Cicero relates the three styles to subject matter: the grand style is suited for ‘lofty’ and ‘grand topics’ ( alta, magna ), the plain style for ‘low’ and ‘small issues’ ( humilia, parva ), and the intermediate style for ‘subjects that range in between’ ( mediocria, modica ). 28 Additionally, he associates the stylistic registers with a triad of discourse functions, or ‘tasks of the orator’ ( officia oratoris ), viz., ‘proving, entertaining and swaying’ ( probare, delectare, flectere ): ‘The plain style is for proof (in probando ), the moderate style for entertainment (in delectando ), and the grand style for swaying ( in flectendo ).’ 29 These three discourse functions should not be confused with the three modes of persuasion, that Cicero expounds in his De Oratore , i.e., ‘teaching, conciliating and moving’ ( docere, conciliare, movere ).30 Furthermore,

27 See esp. Cic. Orat. 70–74. Cicu (2000) discusses Cicero’s adaptation of the term πρέπον into Latin and he explores both the ethical and the aesthetical meanings of the word decorum in Cicero’s rhetorical and philosophical works. 28 Cic. Orat. 100–101. 29 Cic. Orat. 69: subtile in probando, modicum in delectando, vehemens in flectendo. The translation is mine. Cf. Brut. 194–198, Opt. gen. 3. Douglas (1957) notes that the identification of the three styles with three oratorical tasks is a Ciceronian innovation. 30 Both the threefold formula of discourse functions from Brut. , Orat. and Opt. gen. , and the triple division of persuasion modes from De or. are referred to as ‘tasks of the orator’ ( officia oratoris ). The latter triad is articulated in De or. 2.115, 2.121, 2.128–129, 2.310 and 3.104; it can be connected to the Aristotelian πίστεις

107

CHAPTER THREE the three styles can also be related to different parts of the speech: according to Cicero, the orator should use the moderate style in his introductions, he should speak simply and calmly throughout his narrations and argumentations, and he should inflame the passion of the listeners in his perorations. 31 In a word, while Rhetorica ad Herennium simply applies the widely used corporeal metaphor in order to differentiate between the simple and grand style, Cicero adds a moral terminology so as to fit his three styles into the larger framework of his rhetorical program of 46 BC. Dionysius, next, has, as we have seen, two threefold systems instead of one comprehensive triple division (table 2). Dionysius’ ‘types of diction’ ( χαρακτῆρες τῆς λέξεως ) are mostly concerned with selection of words: he describes one style as ‘plain’ or ‘simple’ ( λιτός , ἀφελής , ἰσχνός ), ‘using only the commonest and the most familiar words’, while he characterizes the opposite style as ‘elevated’ ( ὑψηλός ), ‘extraordinary’ (περιττός ) and ‘unusual’ ( ἐξηλλαγμένος ), ‘using archaic, poetical or recondite vocabulary’. 32 The names of the latter style are somewhat ambivalent: whereas the term ὑψηλός has a predominantly positive connotation, περιττός can often be translated in such negative English adjectives as ‘superfluous’, ‘excessive’ or ‘extravagant’. The word ἐξηλλαγμένος , in addition, can mean ‘strange’ and even ‘degenerate’. 33 The choice of terminology reflects Dionysius’ hesitant

from the second book of Rh. (λόγος , ἦθος , πάθος ). Wisse (1989) 212–220 expounds the distinction between the formulas probare, delectare, flectere and docere, conciliare, movere : ‘The division of De Oratore is primarily concerned with content, that in Orator with style and its effect on the audience.’ Sluiter (1994), however, points out that the two divisions are not wholly incompatible. At any rate, Wisse rightly rejects the identification of the three oratorical tasks in De or. with the three styles, as ‘the three styles, which are connected with the three tasks in Orator , only play a minor part in De Oratore , and are not in any way connected with the three pisteis .’ See also Calboli Montefusco (1994), who addresses the difference between the ideas of ‘conciliating’ ( conciliare ) and ‘entertaining’ ( delectare ). Sluiter (1999) 258–260 shows that August. Doctr. christ. 4.32 also distinguishes between the two sets of tasks, one concerning modes of persuasion, the other addressing discourse functions. 31 Cic. Orat. 122. 32 The larger phrases are quoted from Dion. Hal. Dem. 4.1, which discusses the style of Lysianic qualities of the style of Isocrates: Οὔτε γὰρ ἀρχαίοις οὔτε πεποιημένοις οὔτε γλωττηματικοῖς ὀνόμασιν ἀλλὰ τοῖς κοινοτάτοις καὶ συνηθεστάτοις κέχρηται . Cf. ibid. 8, which compares the simple type with the elevated type of diction in similar words. The ambiguity of the term λέξις (both ‘diction’ and ‘style’) has allowed many scholars to interpret Dionysius’ χαρακτῆρες τῆς λέξεως as types of style, that is, not limited to word selection alone: see e.g. Usher (1974) 235–237, Aujac (1988) 16–24, Innes (1989) 269, Wooten (1989) 576 and De Jonge (2008) 180. Yet, Dion. Hal. Dem. 1–34 focuses specifically on selection of words. Cf. n. 24 above on Artés Hernández (2013). 33 For the negative connotations of περιττός and ἐξηλλαγμένος , see LSJ s. περισσός A.II and ibid. s. ἐξαλλάσσω . Both terms carry positive as well as negative connotations in literary criticism. Arist. Pol. 2.3.3, for instance,

108

THE PLAIN , THE GRAND AND THE IN-BETWEEN attitude toward the elevated type of diction: he repeatedly censures the authors who turn to this style for the obscure, far-fetched and unnatural language (section 3.5). He describes the plain style, by contrast, as clear, realistic and resembling natural speech. 34 To be brief, Dionysius principal criterion for distinguishing between his two extreme types of diction is the familiarity of the language. The ‘types of arrangement’ ( χαρακτῆρες τῆς συνθέσεως ) are differentiated according to the acoustic qualities of the word combinations and sentence structure (sections 4.3 and 4.4): on one side of the stylistic spectrum, the arrangement of the words is said to yield a ‘smooth harmony’ ( ἁρμονία γλαφυρά ), while the opposite style is thought to yield a ‘rough harmony’ ( ἁρμονία αὐστηρά ). It should be no surprise that the three types of arrangement do not correspond neatly to the three types of diction, as the two divisions address different aspects of style. The plain type of diction, for instance, does not necessarily correspond to the smooth style of arrangement: the former is described as natural, but the latter is often presented as artificial. Dionysius describes elevated diction, conversely, as unnatural, whereas he presents rough arrangement as simple and unaffected. 35 Still, Dionysius analyzes his three attributes ‘strikingness’ ( τὸ περιττόν ) as a stylistic virtue to Plato’s Socratic dialogues, but Long. Subl. 3.4 connects it to ‘frigidity’ ( ψυχρότης ). The verb ἐξαλλάσσω usually refers to deviations from the common, everyday usage of words and phrase: it is used pejoratively in Arist. Rh. 3.3.3, where it is associated with frigidity ( ψυχρότης ), but Arist. Poet. 22.1458a21–22 connects it with ‘dignified style’ ( λέξις σεμνή ). The perfect participle ἐξηλλαγμένος refers to the result of the procedure of varying common language, i.e., a style with strange and uncommon words and phrases: see e.g. Dion. Hal. Dem. 9.5. 34 De Jonge (2008) 223–240 shows that the concept of the ‘natural’ ( τὸ φυσικόν ) is central throughout Dionysius’ discussions of style: in his later works, the critic ‘develops a more effective way of analyzing the exact character of what he regards as natural (and unnatural) style, syntax and word order’ by turning to a grammatical framework and the method of metathesis. Cf. section 3.3 on the authors that represent the simple (natural) and elevated (unnatural) types of style. In section 3.5, we will see that Dionysius criticizes Plato and Isocrates, two authors whom he associates with the mixed type of diction, especially for their application of the elevated style, while he tends to praise their interpretation of simple diction: cf. esp. Dion. Hal. Dem. 4–7. 35 On the artificiality of smooth arrangement, see Dion. Hal. Comp. 23.3 and 23.7; on the natural appearance of rough arrangement, see ibid. 22.4–5. Pohl (1968) 22–46 expounds the differences between Dionysius’ types of diction and his types of arrangement: see also Bonner (1939) 24, Grube (1974) 78 and Reid (1996) 49–55. In section 3.3 (esp. table 7) below, we will see that Dionysius may associate a classical author with one type of diction, but with a wholly different type of arrangement: Isocrates, for instance, is thought to represent the mixed style of diction and the smooth style of arrangement. Lysias, again, is Dionysius’ paradigm for plain diction, but he does not feature in his discussion of the types of arrangement: Dion. Hal. Lys. 8.4 calls Lysias’ arrangement ‘simple’ ( ἀφελής ) in accordance with his style of diction. Still, there is some overlap between the two divisions: Thucydides, for instance, represents both elevated diction and rough arrangement. Cf. section 3.3 below.

109

CHAPTER THREE types of diction according to similar criteria as his three types of arrangement. The Greek scholar connects both these threefold divisions to his system of stylistic virtues: the table below gives an outline of this system, which distinguishes between so-called ‘essential virtues’ ( ἀναγκαῖαι ἀρεταί ) and ‘additional virtues’ ( ἐπίθετοι ἀρεταί ).36 In his Letter to Pompeius and On Thucydides , Dionysius offers the following division of stylistic virtues in these two categories.37

‘Essential virtues’ ‘Additional virtues’ (ἀναγκαῖαι ἀρεταί ) (ἐπίθετοι ἀρεταί ) 1. Purity of language 1. Vividness (ἐνάργεια ) (τὸ καθαρόν ) 2. Portrayal of characters and emotions (ἠθῶν τε καὶ 2. Clarity (τὸ σαφές ) παθῶν μίμησις ) 3. Brevity (συντομία ) 3. Grandeur (μέγα , μεγαλοπρέπεια ), impressiveness (θαυμαστόν ), sublimity (ὕψος ), beautiful language (καλλιρρημοσύνη ), dignified speech ( σεμνολογία ) 4. Power, forcefulness (ἰσχύς , δεινότης ), intensity (τόνος ), gravity (βάρος ), emotion which rouses the mind (πάθος διεγεῖρον τὸν νοῦν), powerful and combative spirit (ἐρρωμένον καὶ ἐναγώνιον ῥεῦμα ) 5. Persuasion ( πειθώ ), pleasure (ἡδονή ), delight ( τέρψις ), charm ( χάρις ) 6. Appropriateness ( τὸ πρέπον )

Table 4: Dionysius’ system of stylistic virtues (Τhuc. 23.6–7, Pomp. 3.8–10)

36 For discussions of Dionysius’ system of stylistic virtues and its relationship with other (earlier) theories about stylistic virtues in Aristotle, Theophrastus and Stoic sources, see Stroux (1912) 72–88, Meerwaldt (1920) 3–26 and Bonner (1939) 16–20. Dionysius’ complex system, then, seems to have been a later development in the history of rhetoric, but it is unlikely that it was his own invention: at Thuc. 22.2, Dionysius submits that the division ‘has been laid out by many authors before’ ( εἴρηται πολλοῖς πρότερον ). On several occasions, Cicero gives expression to the view that all oratory should exhibit certain essential virtues (e.g. pure Latin, clarity), while praise and admiration are reserved for those orators who master the more complex, additional virtues (e.g. rhythm, appropriateness): see Part. or. 31–32, De or. 3.38, 3.52–54, Brut. 261. Thus, Cicero may have been aware of a system of necessary and additional virtues, but he does not offer a full-blown exposition. 37 See Dion. Hal. Thuc. 22–23 and Pomp. 3.8–10. It is the Greek critic himself who suggests the division of additional virtues into the six subcategories presented in table 4.

110

THE PLAIN , THE GRAND AND THE IN-BETWEEN

In what way does this complex system of stylistic virtues contribute to our understanding of Dionysius’ types of diction and types of arrangement? The attributes that the critic ascribes to simple and to elevated diction, as well as those that he ascribes to smooth and rough arrangement can be divided into a necessary and an additional category. He associates simple diction, for one, with ‘what is necessary and useful’ ( τἀναγκαῖα καὶ χρήσιμα ), while he often refers to the virtues of elevated diction as ‘additional ornaments’ ( ἐπίθετοι κόσμοι ). 38 The table below presents a (non-exhaustive) list of attributes that Dionysius applies to both extreme styles of word selection. The two groups of adjectives below correspond roughly, though not perfectly, with Dionysius’ division between necessary and additional virtues, that we have seen above (table 4). 39

Attributes of simple diction Attributes of elevated diction Simple ( λιτός , ἁπλοῦς, ἀφελής ) Elevated ( ὑψηλός ) Ordinary (ἀπέριττος , συνήθης , κοινός ) Unusual ( περιττός , ἐξηλλαγμένος , ξένος ) Sweet (ἡδύς ) Bitter ( πικρός ) Cheerful ( ἱλαρός ) Solemn ( σεμνός ) Uncontrived ( ἀποίητος , ἀνεπιτήδευτος ) Artificial ( ἐγκατάσκευος , ἐπιτήδειος ) Real, practical ( ἀληθινός ) Ceremonial (πανηγυρικός ) Safe ( ἀσφαλής , ἀκίνδυνος ) Daring, risky ( τολμηρός ) Relaxed (ἀνειμένος ) Intense ( σύντονος ) Displaying moral character (ἠθικός ) Arousing emotion (παθητικός ) Pure ( καθαρός ) Weighty ( βαρύς ) Clear ( σαφής ) Rough ( αὐστηρός ) Concise ( σύντομος ) Beautiful ( καλός ) Persuasive ( πιθανός ) Dignified ( ἀξιωματικός ) Accurate (ἀκριβής ) Grand ( μεγαλοπρεπής )

Table 5: attributes of simple and elevated diction (Dion. Hal. Dem. 1–34)

38 Dion. Hal. Dem. 3.2. Cf. ibid. 1.3, 13.7 and 18.1. 39 The virtues can be found throughout Dem. 1–34, the first half of the treatise. For the sake of uniformity I have included only adjectives in the table, although one can easily find related nouns in Dem. (e.g. μεγαλοπρέπεια , τόνος ): I have used the nominative singular masculine forms of the adjectives, regardless of the forms in which they actually appear in Dionysius’ text. Concerning the apparent differences between tables 4 and 5, cf. Porter (2016) 221 and section 1.5 n. 72 above on the ‘provisional nature’ of Dionysius’ theoretical schemas.

111

CHAPTER THREE

Dionysius has a similar division of virtues in his discussion of the types of arrangement. In On the Arrangement of Words , he recognizes two main virtues to differentiate between the opposite styles of word arrangement, ‘pleasure’ ( ἡδονή ) for smooth harmony and ‘beauty’ (κάλλος ) for rough harmony. The Greek author subsequently offers two concise lists of virtues that fall under these two categories. 40 As we can see in the table below, the division quite neatly (albeit, again, not completely) matches Dionysius’ system of necessary and additional virtues (table 4) and his description of simple and elevated diction (table 5).

Attributes of smooth arrangement Attributes of rough arrangement Pleasure ( ἡδονή ) Beauty ( κάλλος ) Freshness ( ὥρα ) Weight (βάρος ) Charm ( χάρις ) Grandeur (μεγαλοπρέπεια ) Euphony ( εὐστομία ) Solemnity (σεμνολογία ) Sweetness ( γλυκύτης ) Dignity (ἀξίωμα ) Persuasiveness ( τὸ πιθανόν ) Emotion (πάθος )

Table 6: attributes of smooth and rough arrangement (Dion. Hal. Comp. 11.2–3)

Simple diction is characterized by the same sort of virtues as smooth harmony, and elevated diction is described in a similar vocabulary as rough harmony. Like Cicero, Dionysius fits his views on stylistic registers into the larger framework of his literary doctrines: while the Roman rhetorician had connected his discussion of the three styles to his theory of the orator’s tasks, the Greek critic links each of his threefold divisions to his theory of stylistic virtues. The malleability of the three-style doctrine allows our authors to adapt it so as to incorporate it into the overarching stylistic program of their rhetorical and critical treatises.

3.2.2 The Flexibility of the Middle Style In the previous subsection, we have seen that our sources use various antitheses as guiding principles for their stylistic divisions—thin vs. heavy, calm vs. passionate, ordinary vs.

40 See Dion. Hal. Comp. 10.1–2 for the virtues of beauty and pleasure, and ibid. 11.2–3 for the virtues associated with the two cardinal virtues. The opposition between ‘beauty’ ( κάλλος , τὸ καλόν ) and ‘pleasure’ ( ἡδονή , τὸ ἡδύ ) is essential in Dionysius’ critical treatises, esp. in his discussions of word arrangement. Donadi (1986) argues that the distinction is a Dionysian innovation, while Goudriaan (1989) 203–215 connects the division to the author’s views on politics and civilization. Cf. section 2.3.3 n. 84 above and esp. section 4.5 below on the role of beauty and pleasure in Dionysius’ evaluation of word arrangement.

112

THE PLAIN , THE GRAND AND THE IN-BETWEEN unusual, smooth vs. rough. The present subsection will focus on the extant interpretations of the intervening type of style that is somehow situated between these extremes. In Latin sources, this intermediate style holds a middle between the two extremes, but the rhetoricians do not explain how exactly this middle ground is related to the other two styles. According to the author of Rhetorica ad Herennium , the intermediate style ‘consists of a lower (viz., than the grand style), yet not of the lowest and most colloquial class of words’. 41 The author highlights the indefinite character of this style, when he introduces the vicious neighboring style, which he calls the ‘slack style’ ( figura dissoluta ) and which ‘drifts to and fro’ ( fluctuat huc et illuc ), being utterly ‘without sinews and joints’ ( sine nervis et articulis ). 42 Precisely because the mean is such a vague concept, the adept of the middle style runs the risk of composing a vague discourse. Cicero agrees with the author of Rhetorica ad Herennium about the vagueness of the middle style: ‘It is akin to both other style, excelling in neither, sharing in both, or, to tell the truth, sharing in neither.’ 43 According to Cicero, it does not possess the ‘sharp edge’ ( acumen ) of the simple style nor the forceful ‘thunder’ ( fulmen ) of the grand style: consequently, it has an ‘absolute minimum of vigor’ ( nervorum vel minimum ). Yet, Cicero submits that this floating style has its own merits: it makes up for its lack of force by supplying an ‘absolute maximum of charm’ ( suavitatis vel plurimum ). 44 In his Orator , Cicero associates the middle style particularly with the relaxed joy of epideictic oratory: in his view, the middle style is not

41 Rhet. Her. 4.11: Constat ex humiliore neque tamen ex infima et pervulgatissima verborum dignitate. Cf. ibid. 4.13, which considers any given discourse to belong to the middle style, ‘if we have somewhat relaxed our style, and yet have not descended to the most ordinary prose’ ( si haec aliquantum demiserimus neque tamen ad infimum descenderimus ). 42 Rhet. Her. 4.16. Cf. Gell. NA 6.14.5, who calls the orators who are unsuccessful in composing a discourse of the middle style ‘unclear and ambiguous’ ( incerti et ambigui ). Cf. section 3.2.1, esp. n. 26 above on the faulty styles in Rhet. Her. , Demetr. Eloc. and Gell. NA. For the phrase ‘without sinews’ ( sine nervis ), cf. Cic. Orat. 91. See also the description of the middle style in Rhet. Her. by Leeman (1963) 30: it ‘is mainly characterised by what it is not. It does not have the pathos of the grandis figura , nor the naked simplicity of the figura attenuata . What remains is a pleasant and relaxed loftiness.’ The notions of pleasure and relaxation accord well with Cicero’s interpretation of the middle style: see the next paragraph below. 43 Cic. Orat. 21: Vicinus amborum, in neutro excellens, utriusque particeps vel utriusque, si verum quaerimus, potius expers. Cf. Winterbottom (1989) 127: the middle style is a ‘colourless half-way house’. 44 See Cic. Orat. 21 ( acumen, fulmen ) and ibid. 91: ‘In this style there is perhaps a minimum of vigor and a maximum of charm’ ( hoc in genere nervorum vel minimum, suavitatis autem est vel plurmimum ). Cf. Rhet. Her. 4.16: ‘Without sinews and joints’ ( sine nervis et articulis ). On lightning as a productive symbol for grand oratory in general and for Demosthenes in particular in Cicero’s works, see section 2.3.2 above.

113

CHAPTER THREE indigenous to the Forum, but ‘the sophists are the source ( e sophistarum fontibus ) from which all this has flowed into the Forum ( defluxit in forum ), but scorned by the simple and rejected by the grand, it found a resting-place in this middle class ( in mediocritate ) of which I am speaking’. 45 Cicero presents the pleasant oratory of Demetrius of Phalerum as the paradigm of the intermediate style: the Peripatetic scholar ‘came forth in the sun and dust of action, not as from a soldier’s tent, but from the shady retreat (ex umbraculis ) of the great philosopher Theophrastus’ (section 3.3).46 Dionysius has a rather different conception of the intermediate style than his Roman predecessors: on some occasions, the Greek critic describes it, like Cicero and the author of Rhetorica ad Herennium , as ‘intermediary’ ( μέσος , μεταξύ ) between the two extreme styles, but at other times he proposes that it is a ‘mixture’ ( μῖγμα ), a ‘blend’ ( κρᾶσις ) or a ‘combination’ (σύνθεσις ) of them. As we will see, Dionysius’ middle style is actually nothing more than an appropriate application of the extreme styles, that is, of both simple and elevated diction, of both smooth and rough arrangement—in a word, of both the essential and the additional virtues. Indeed, the middle style ‘has no form peculiar to itself’. 47 Dionysius’ interpretation of the stylistic mean merits a closer look, not only because of the apparently

45 Cic. Orat. 96: Hoc totum e sophistarum fontibus defluxit in forum, sed spretum a subtilibus, repulsum a gravibus in ea de qua loquor mediocritate consedit. Cicero draws a contrast between the practical force of the simple and grand style on the one hand and the relaxed entertainment associated with the middle style: cf. section 2.3.1 n. 44 above on the frequently attested association of practical oratory with military weapons and the bodies of soldiers, and of epideictic oratory with ceremonial weapons and the bodies of athletes. Winterbottom (1989) argues that Cicero’s conflation of the middle style with epideictic oratory serves as an apology for the epideictic elements in his own oratory. The connections between Cicero’s middle style and epideictic oratory are expounded by Von Albrecht (2003) 19–20. 46 Cic. Brut. 37: Processerat enim in solem et pulverem, non ut e militari tabernaculo, sed ut e Theophrasti doctissimi hominis umbraculis. For Demetrius of Phalerum, see section 1.6 n. 128 above (on the transition from Attic to Asian oratory) and section 3.3 below (on Demetrius’ middle style). Quint. Inst. orat. 12.10.58 may have Cicero’s discussion in mind in his description of the middle style as ‘flowery’ ( genus floridum ). 47 Dion. Hal. Dem. 41.1: ‘There is no quality peculiar to itself’ ( οὐδείς ἐστι χαρακτὴρ ἴδιος ). Dion. Hal. Comp. 24.1: ‘It has no form peculiar to itself’ ( σχῆμα ἴδιον οὐδὲν ἔχει ). On the necessary and additional virtues in the mixed style of diction, see Dem. 3.2. Martinho (2010) studies the terminology that Dionysius uses for the middle style, focusing particularly on the concept of mixture. Yunis (2019) refers to this concept as Dionysius’ ‘most interesting and effective idea for rhetorical style’. Cf. section 3.5 below for the two different forms of mixture (μῖγμα ) that Dionysius introduces in his works, that is, blending ( κρᾶσις ) and combining ( σύνθεσις ). Note that the idea of mixing appears rarely in other critical texts: Demetr. Eloc. 36 ( μίγνυται ) and Cic. Orat. 21 (temperatus , cf. OLD s. tempero, ‘mingle in due proportion’) briefly touch on it, without further elaboration.

114

THE PLAIN , THE GRAND AND THE IN-BETWEEN ambiguous terminology of mixing, blending and combining, but even more so because Dionysius identifies the middle style as the best possible style, not a bland mean, but rather a formidable blend of the best qualities of the other styles. In a later part of this chapter, I will therefore address Dionysius’ peculiar conception of the middle and his reasons for proclaiming it the best (section 3.5). Thus, the Greek and Latin sources do not only exhibit various approaches to the two binary opposed styles that form the basis of their threefold divisions, but their views also differ considerably with respect to the intermediate style, which lies between the extremities. While the author of Rhetorica ad Herennium and Cicero portray it as a vague mean which lacks vigor and which, in Cicero’s view, is mainly designed to entertain, Dionysius’ middle style unites all the best features of the other two styles. To the Greek scholar, the mixed style, far from being weak, is the ‘most powerful’ ( κράτιστος ) of the three available registers. 48

3.3 Classicism and the Three Styles The theory of three styles is closely connected to the classicizing approach to style that was in vogue in the first century BC: as Roman and Greek authors alike studied the literary models of classical Greece with increasing intensity, the theory of three styles came to be used as a convenient, versatile tool to analyze, categorize and compare the classics. 49 Although the theory of three styles was especially influential in rhetorical theory, other genres could also be studied within its framework: Dionysius applies his three types of word arrangement, for instance, to epic, lyric, tragedy as well as to rhetoric, history and philosophy. 50 Furthermore, the three styles could not only be applied to Greek but also to Latin literature: we know that the Roman polymath Varro proposed Latin examples for each of the three styles: he attributed a grand style to the tragic poet Pacuvius, an intermediate style to the comic playwright Terence, and a plain style to the satirist Lucilius.51 In this section, I will focus on the

48 See Dion. Hal. Dem. 15.1 and 34.5 on the superiority of the middle type of diction, and Comp. 24.2 on Dionysius’ view that the middle type of arrangement carries off the ‘first prize’ ( τὰ πρωτεῖα). 49 For an introduction to classicism and the shared approach to classical Greek literature by Greek and Roman authors in Rome, see section 1.6 above and the secondary literature cited there. 50 Cf. n. 60 below for the Dionysius’ division of ancient Greek poetry according to his three types of word arrangement. See also Gell. NA 6.14.1: ‘Both in verse and in prose there are three approved styles’ ( et in carmine et in soluta oratione genera dicendi probabilia sunt tria ). 51 Gell. NA 6.14.6 (= Varro fr. 80 Wilmanns) credits Varro with the view that the tragedies of Pacuvius displays ‘abundance’ (ubertas ), the comedies of Terence a ‘middle style’ ( mediocritas ), and the satires of Lucilius ‘slenderness’ ( gracilitas ). These associations are clearly made on account of the genres rather than the styles that

115

CHAPTER THREE application of the three-style doctrine in the analysis of classical Greek prose: I will specifically compare the lists of authors that Cicero and Dionysius supply as exemplary models for each of the registers in their stylistic divisions. First of all, however, it is worthwhile to notice that the critics and rhetoricians in Rome traced the invention of the triple stylistic division itself back to the classical past: various Greek and Latin authors connected the three styles to pivotal moments in the history of both Greek and Latin rhetoric. Many considered Homer’s Iliad to be such a watershed: the old poet was not only generally held to have been the founding father of rhetoric in general, but he was also thought to have been the first to formulate the theory of three styles. 52 Indeed, various sources state that Homer assigned a plain style to Menelaus, an elevated style to Odysseus, and an intermediate style to Nestor. 53 The poet described Menelaus, for one, as speaking ‘concisely but very clearly’ ( παῦρα μὲν ἀλλὰ μάλα λιγέως ): ‘He was not a man of many words, nor did he not speak at random’ ( οὐ πολύμυθος οὐδ’ ἀφαμαρτοεπής ). Odysseus, conversely, was impressive: ‘His words were like snowflakes on a winter’s day’ ( ἔπεα are peculiar to the authors themselves: tragedy is grand, satire is simple, and comedy holds a mean between the former two genres. It is by no means certain that Varro ever applied the Latin terms ubertas, gracilitas and mediocritas . At any rate, Gellius does not attribute his entire discussion of the three styles to Varro: the latter is not mentioned, for instance, when Gellius introduces the Greek and Latin names for the three styles, viz., ἁδρός /uber (‘rich’), ἰσχνός /gracilis (‘thin’) and μέσος /mediocris (‘intermediate’), and the corresponding erroneous styles (cf. n. 26 above). Grube (1965) 163 notes that, ‘if Gellius is quoting correctly, Varro’s was the first extant example of the use of the three-style formula’. Cf. section 3.1 n. 1 above. 52 According to Sluiter (2005) 392–396, the rhetoricians’ interest in Homer’s epics was threefold: (1) rhetoric could be ‘found back’ in Homer, for which the attribution of the three styles to three Homeric heroes is a clear example, (2) rhetorical analysis could be applied to the exegesis of Homeric passages, and (3) Homeric rhetoric could be used as textbook material in the education of orators. Recently, Ahern Knudsen (2014) 1–14 has argued that ‘a latent theory of rhetoric exists in Homer’: she recognizes thirty-five ‘rhetorical speeches’ in Hom. Il. , arguing that they display techniques of persuasion similar to the ones advocated in Arist. Rh. 53 Cf. section 3.1 n. 7 above. Radermacher (1951) 6–9 gives a list of sources that connect the three styles to Homer: see esp. Cic. Brut. 40, 50; Laus Pisonis 57–64; Sen. Ep. 40.2; Quint. Inst. orat. 12.10.64; Gell. NA 6.14.5; Fronto De eloc. 1.5; [Plut.] Vit. Hom. 172; Hermog. Id. 2.9 Rabe; Aristid. In Plat. de rhet. p. 30–31 Dindorf; Eust. Il. 3.212. Cf. Nünlist (2009) 220 on schol. AbT Il. 3.212. Although Cicero is the earliest extant source for the attribution of stylistic registers to Homeric heroes, it is often suggested that the topic of the three styles of rhetoric in Homer goes back to at least the fourth century BC: Yet, Kennedy (1957a) 26–29 rightly notes that references to Homeric rhetoric in Pl. Phdr. 261b and Antisthenes (ap. Porph. Schol. Od. 1.1) do not refer to stylistic divisions: ‘The topic is, therefore, primarily one of the Hellenistic age, and it is to that period and its schools that we must now turn.’ Given the provenance of the surviving texts, I submit that the issue was primarily one of Greek and Roman critics in Rome, as was the whole three-style division: cf. section 3.1.

116

THE PLAIN , THE GRAND AND THE IN-BETWEEN

νιφάδεσσιν ἐοικότα χειμερίῃσιν ). Nestor’s style, lastly, was ‘pleasant’ ( ἡδυεπής ): from his lips came forth a voice ‘sweeter than honey’ ( μέλιτος γλυκίων ). 54 Homer, then, was widely held to have been the inventor of the theory of three styles. The threefold doctrine, in addition, was not only associated with the birth of rhetoric in Greece, but also with its arrival in Rome. In 155 BC, an Athenian embassy visited Rome to negotiate with the senate about the reduction of a fine that had been imposed on the Athenians. 55 Among the envoys were three famous philosophers, Carneades from the Academy, Critolaus the Peripatetic and Diogenes the Stoic, who beside their official duties in the senate also gave public displays of their eloquence. According to Aulus Gellius, the philosophers each represented one of the three canonical styles: ‘Carneades spoke with a vehemence that carried you away ( violenta et rapida ), Critolaus with art and polish ( scita et teretia ), Diogenes with restraint and sobriety ( modesta et sobria ).’ 56 Thus, the Platonist stands for the elevated style, the Aristotelian for the middle style, and the Stoic for the plain style. 57

54 See Hom. Il. 3.212–224 (Menelaus and Odysseus) and ibid. 1.247–249 (Nestor); translations are mine. 55 The embassy was famous in Antiquity because of the rhetorical displays of the participating Greek philosophers and the ambivalent reactions that they evoked from the Romans: see esp. Powell (2013b), who discusses the sources and concludes that ‘in reality the embassy may not have been such an iconic moment for the Roman encounter with Greek philosophy as it is usually supposed to have been, until Cicero made it so’. The main ancient sources are Cic. De or. 2.155–161, Plut. Cat. Mai. 22, Plin. NH 7.112, Quint. Inst. orat. 12.1.35, Gell. NA 6.14.8–10, Lactant. Div. inst. 5.14.3–5. In fact, Greek and Roman authors associated various supposed key moments in the history of rhetoric and grammar with foreign embassies. Gorgias’ visit to Athens in 427 BC as an envoy of his native town Leontini, for instance, was sometimes considered to have been the starting point of artificial prose in Athens: see e.g. Diod. Sic. 12.53.4 and Dion. Hal. Lys. 3.3 (= Timaeus F137 Jacoby). For Gorgias as the inventor of artificial prose, see Cic. Orat. 39–40, 165, 175. Likewise, Suet. Gramm. 2.1–4 reports that Crates of Mallos gave the first impulse to the study of grammar and criticism in Rome, while he visited the city as an envoy of the Attalid court in 168 BC: see section 4.2, esp. n. 18 below. 56 Gell. NA 6.14.8–10: Violenta et rapida Carneades dicebat, scita et teretia Critolaus, modesta Diogenes et sobria . The words that Gellius uses do not appear in Rhet. Her., Cic. or Quint. Inst. orat. to describe the three styles: hence, Gellius supplies another illustration of the versatility of the theory of three styles. The author’s claim that Polybius and Rutilius Rufus are his sources is very dubious: see n. 58 below. 57 The distribution of the three styles over the three philosophers tallies with the ancient reputations of Academy, Peripatos and Stoa: Plato, for one, is often associated with the elevated style (cf. e.g. Dion. Hal. Dem. 5.4–6, Long. Subl. 4.6), while Aristotle and his followers famously advocated the mean ( μεσότης ) as the seat of virtue (cf., e.g., Dion. Hal. Comp. 24.2, section 3.5 below), and Stoic philosophers propagated a simple, concise type of rhetoric: cf. Moretti (1990). Diog. Laert. 7.59 credits the Stoic Diogenes of Babylon, one of the visitors of Rome in 155 BC, with the introduction of ‘brevity’ ( συντομία ) as the fifth cardinal stylistic virtue, in addition to correctness of language, clarity, appropriateness and ornamentation.

117

CHAPTER THREE

Gellius claims that he took the anecdote from Polybius and Rutilius Rufus, but it seems equally plausible that his source was the lost third book of Cicero’s dialogue On the Republic , which also discussed the embassy. 58 In any case, the three styles are once again connected to an iconic moment in the history of rhetoric.

Cicero Dionysius

genera dicendi (e.g. Brut. χαρακτῆρες τῆς λέξεως χαρακτῆρες τῆς συνθέσεως 67–88, Orat. 15, 92, 111) (Dem. 1–3) (Comp. 22.7, 23.9, 24.5)

Isocrates, Ephorus, 1 Lysias, Hyperides Lysias Theopompus

Thrasymachus, Isocrates, Demosthenes, Herodotus, 2 Demetrius of Phalerum Plato, Demosthenes Democritus, Plato, Aristotle

3 Pericles, Demosthenes Thucydides, Gorgias Antiphon, Thucydides

Table 7: Cicero’s and Dionysius’ classifications of classical Greek prose according to the three styles

Needless to say, the theory of three styles was not primarily designed to compare Homeric heroes and philosophical delegates: it was more appositely put to use to analyze the divergent styles exhibited in the corpus of classical Greek literature. The table above lists the Greek prose authors that Cicero and Dionysius consider typical champions for each of the three styles. Naturally, Dionysius gives us two separate lists, one for the types of diction (simple, mixed and elevated respectively) and another for the types of arrangement (smooth, mixed and rough respectively): I have already explored the differences between these two divisions above (section 3.2.1). Cicero does not offer a comprehensive canon of authors: his references

58 Powell (2013) 243–244 points out that the third book of Cic. Rep. mentions Rutilius Rufus as a fictional intermediary through whom the dialogue was transmitted, and that Polybius is also mentioned in the same book: ‘The mention of the two names together makes one wonder whether Gellius would have looked up both those sources separately and whether they both said the same thing, or whether he used one source which happened to mention them both.’ It is not unlikely, surely, that the topic of rhetorical styles was of interest to Cicero in the period between 54 and 51 BC, when he was working on Rep. : see Zetzel (1995) 1–3.

118

THE PLAIN , THE GRAND AND THE IN-BETWEEN to the authors and their styles are scattered over his three rhetorical works of 46 BC. Note that Rhetorica ad Herennium does not feature in the overview, as it does not connect the three styles to specific Greek or Latin authors: the anonymous rhetorician, as we have seen, rejects the ‘practice of Greek writers’ ( consuetudo Graecorum ) of drawing stylistic examples from famous orators or poets. 59 I have also omitted from the table above the poets that Dionysius connects to his three types of arrangement, as they cannot readily be compared to Cicero’s list, which only refers to orators. 60 In the simple style, both Cicero and Dionysius select Lysias as the principal model. In the same category, Dionysius inscribes ‘the genealogists, those who dealt with local history, the natural philosophers and the moral philosophers, including the entire Socratic school (except for Plato) and almost all those who composed political and forensic speeches’. 61 Dionysius does not connect Lysias to any of his three types of harmony: he describes the orator’s word arrangement as simple and seemingly spontaneous.62 Cicero and Dionysius, to continue, select different authors as antipodes to Lysias’ plain style. Cicero, for one, contrasts him with Pericles and Demosthenes, who were both thought, as we saw before, to display a thunderlike quality (section 2.3.2): Cicero repeats Aristophanes’ statement that Pericles ‘lightened, and thundered, and embroiled the whole of Greece’, while he often refers to the ‘thunderbolts’ ( fulmina ) of Demosthenes. 63 Dionysius, however, pits the simple word choice

59 See Rhet. Her. 4.1–10: cf. section 1.4 above. On the author’s self-written samples of the three styles and the neighboring vicious styles, see section 3.2.1 above, esp. n. 26. 60 Dion. Hal. Comp. 22.7, 23.9 and 24.5 gives the following division of poets according to the three styles. Among the epic poets, Dionysius presents Antimachus and Empedocles as rough, Hesiod as smooth, and Homer is mixed. As for the lyric poets, he classifies Pindar as rough, Sappho, Anacreon and Simonides as smooth, and Stesichorus and Alcaeus as mixed. In tragedy, lastly, he puts forward Aeschylus as rough, Euripides as smooth, and Sophocles as mixed. 61 Dion. Hal. Dem. 2.2: Οἱ γενεαλογίας ἐξενέγκαντες καὶ οἱ τὰς τοπικὰς ἱστορίας πραγματευσάμενοι καὶ οἱ τὰ φυσικὰ φιλοσοφήσαντες καὶ οἱ τῶν ἠθικῶν διαλόγων ποιηταί , ὧν ἦν τὸ Σωκρατικὸν διδασκαλεῖον πᾶν ἔξω Πλάτωνος , καὶ οἱ τοὺς δημγορικοὺς ἢ δικανικοὺς συνταττόμενοι λόγους ὀλίγου δεῖν πάντες . 62 See Dion. Hal. Lys. 8.2, describing Lysias’ arrangement as ‘absolutely simple and plain’ ( ἀφελὴς πάνυ καὶ ἁπλοῦς). This description seems compatible with Dionysius’ conception of rough harmony: at Comp. 22.5, the critic describes rough arrangement as ‘artless’ ( ἀνεπιτήδευτος ) and ‘simple’ ( ἀφελής ), cf. also section 4.5 below. Note, however, that Dion. Hal. Din. 6.4 connects Lysias’ naturalness to smoothness, not roughness: his composition is ‘apparently natural and smooth’ ( αὐτοφυὴς καὶ λεία εἶναι δοκοῦσα ). It seems, then, that Lysias’ style of arrangement cannot be easily fitted into Dionysius threefold system: cf. section 3.2.1 n. 35 above. 63 For the frequently attested connections between thunderbolts and sublime literature, see section 2.3.2 n. 59. Cf. also ibid. for the ancient sources that link Pericles’ oratory to thunder and lightning. The passage from

119

CHAPTER THREE of Lysias against the elevated diction of Gorgias and Thucydides. Together with the orator Antiphon, Thucydides is also Dionysius’ champion of rough arrangement, standing in opposition to the smooth harmony of Isocrates and his pupils Theopompus and Ephorus. 64 Thucydides, whose history is considered exemplary for elevated vocabulary and rough arrangement, is considered a typical adept of the additional virtues of style (section 3.2.1). The authors that Cicero lists are all orators, whereas Dionysius includes historians (Theopompus, Ephorus, Herodotus, Thucydides) and philosophers (Democritus, Plato, Aristotle) as well as poets. The omission of all other prose authors except for orators in Cicero’s discussion of the three styles is not arbitrary: the Roman rhetorician emphasizes that the styles of historiography and philosophy have no place on the Forum. The style of the philosophers, in his view, ‘lacks the vigor and the sting necessary for oratorical efforts in public life’. 65 The historians, in addition, aim to achieve a ‘calm and flowing style, not the terse and vigorous language of the Forum’. Apparently, Cicero deems the style of these two genres unfit for practical, everyday situations in the law-courts and the assemblies (section 2.3.4 and 2.4.4). Nonetheless, in the case of historiography, I think that Cicero would not be unhappy with Dionysius’ division. Cicero calls Ephorus, for example, ‘smooth’ ( levis ), and he assigns to Theopompus a periodic style without hiatus, which tallies well with Dionysius’ description of the smooth type of arrangement. Herodotus, in Cicero’s view, ‘flows along like a peaceful stream without any rough waters’, whereas Thucydides ‘moves with greater vigor, and in his description of war, he sounds, as it were, the trumpet of war’. 66

Aristophanes is Ach. 530–531: ‘And then in wrath Pericles, that Olympian, did lighten and thunder and stir up Greece’ ( ἐντεῦθεν ὀργῇ Περικλέης οὑλύμπιος / ἤστραπτ ’, ἐβρόντα , ξυνεκύκα τὴν Ἑλλάδα ). According to Cic. Orat. 29, Aristophanes said ‘that Pericles lightened, and thundered, and embroiled all Greece’ ( fulgere, tonare permiscere Graeciam ). 64 Cicero, too, associates Isocrates’ school with smoothness. Cic. Orat. 151 attributes to Isocrates and Theopompus, for instance, the practice of avoiding hiatus, which is mentioned as a trademark of smooth word arrangement by Dion. Hal. Comp. 23.13. Cic. Orat. 191 calls Ephorus ‘smooth’ ( levis ). The Roman rhetorician, moreover, associates Isocrates, Ephorus and Theopompus repeatedly with carefully structured rhythmical periods: see e.g. Orat. 151, 172, 174, 191, 207, 218. Cf. also the next paragraph below. 65 Cic. Orat. 62: Horum oratio neque nervos neque aculeos oratorios ac forensis habet. Ibid. 66: In his tracta quaedam et fluens expetitur, non haec contorta et acris oratio. Note that Cicero submits that some philosophers, specifically Plato, Aristotle, Xenophon and Theophrastus wrote ‘ornately’ ( ornate ) and could easily challenge the orators of the courts and the assemblies. For Dionysius’ comparison of Plato’s and Demosthenes’ styles, see section 3.5 below. 66 On Herodotus and Thucydides, see Cic. Orat. 39: Alter enim sine ullis salebris quasi sedatus amnis fluit, alter incitatior fertur et de bellicis rebus canit etiam quodam modo bellicum. The pleasantness of Herodotus’ sedatus

120

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Two names on Cicero’s and Dionysius’ lists deserve closer attention. The first is Demosthenes, who occupies a special place in all threefold divisions. As we saw, Cicero and Dionysius both think him superior to the other prose authors of Classical Greece, in any type of style (section 2.3.2–2.3.4). According to the Roman rhetorician, ‘nobody is more impressive ( gravior ), nobody more astute ( callidior ), nobody more tempered ( temperatior ) than him’. Similarly, the Greek critic claims that Demosthenes was ‘more successful than the other authors’ in all three styles. 67 Thus, Cicero and Dionysius agree that Demosthenes does not pretend to any single style. For Dionysius, this makes the orator the consummate icon of the mixed style of diction and arrangement, which, after all, is nothing else than an appropriate combination of the best qualities of the other two styles. According to Cicero, however, Demosthenes is at his best in the grand style: ‘He gets the applause and makes his speech count for the most, when he uses the topics of the impressive style ( gravitatis loci ).’ 68 In a word, Cicero and Dionysius both stress the importance of stylistic versatility, but they articulate their ideals differently: the younger Greek scholar insists that the mixed style is superior, whereas the elder Roman prioritizes the grand style. The second name that merits a moment of our attention is Demetrius of Phalerum (fl. ca. 300 BC), whom Cicero selects as the prime exponent of the intermediate style. As noted above, however, he was a quite ambivalent figure, as he was active at the time of the alleged decline of Attic oratory and the concomitant rise of its Asian counterpart (section 1.6). In De Oratore , Cicero portrays him as the initiator of ‘softer and more relaxed types of style’ (molliora ac remissiora genera ), which Cicero recognized in the famous, or infamous, Asian orators Menecles and Hierocles of Alabanda. 69 Nevertheless, Cicero admires the calm and

amnis can easily be connected to Cicero’s middle style, which, after all, aims at ‘charm’ ( suavitas ) and ‘entertainment’ ( delectare ). The epithet incitatus , that Cicero attaches to Thucydides, is akin to the grand, or ‘passionate’ ( vehemens ), style. For Cicero’s discussion of the styles of Theopompus and Ephorus, see n. 64 above. 67 Cic. Orat. 23: Hoc nec gravior exstitit quisquam nec callidior nec temperatior. Dion. Hal. Dem. 33.3 describes Demosthenes as ‘being the most successful of all the other authors in the three types of style’ ( ἐν τοῖς τρισὶ γένεσι κατορθῶν τῶν ἄλλων μάλιστα ). 68 Cic. Orat. 111: Clamores tamen tum movet et tum in dicendo plurimum efficit cum gravitatis locis utitur. 69 Cic. De or. 2.94–95 declares Lycurgus, Demosthenes, Hyperides, Aeschines and Dinarchus to have been the last generation of truly Attic orators, after whose deaths started the habit of writing ‘softer and more relaxed styles’ ( molliora ac remissiora genera ), which in Cicero’s own day could be found in the speeches of Menecles and Hierocles: see section 1.6 n. 128 above. For Demetrius as a transitional figure between Attic and Asian oratory, see Cic. Brut. 37–38 (= Demetrius of Phalerum fr. 175 Wehrli): ‘He was the first to modulate oratory

121

CHAPTER THREE peaceful flow of Demetrius’ rhetoric, which makes it particularly apt for providing pleasure to the audience: hence, ‘he entertained rather than stirred his countrymen’. Much like the middle style itself, as we have seen, Demetrius’ oratory is thought not to be particularly at home in the struggles of the Forum (section 3.2.2): ‘His training was less for the field than for the parade-ground’ and ‘he chose to use charm rather than force, a charm which diffused itself through the minds of his listeners without overwhelming them’. 70 Thus, Demetrius represents the pleasant vagueness that is the hallmark of Cicero’s intermediate style. To summarize, the threefold divisions of Cicero and Dionysius yield divergent classical canons. In the next two sections, I will zoom in on the goals and motivations that underlie the author lists of the Roman rhetorician and the Greek critic. Cicero’s canon can to a large extent be ascribed to his opposition to the claims of the so-called Attici (section 3.4). Dionysius’ canon, next, reflects the author’s insistence on versatile eloquence with universal appeal (section 3.5).

3.4 Cicero against the So-Called Attici In the decennium between 55 and 45 BC, as we saw in the introduction to this dissertation (section 1.4), a group of Roman orators who called themselves ‘Attic’ ( Attici ) rose to prominence on the Forum. 71 Presumably led by the orator-cum-poet C. Licinius Macer Calvus (whose two cognomina, incidentally, mean ‘meager’ and ‘bald’ respectively), these ‘Neo- and give it softness and pliability, and he chose to use charm, as was his nature, rather than force’ ( hic primus inflexit orationem et eam mollem teneramque reddidit et suavis, sicut fuit, videri maluit ). Douglas (1966) 28 and (1973) 95 discusses three possible interpretations of the phrase ‘modulating oratory’ ( inflectere orationem ); on the basis of the elder Seneca’s treatment of the declaimer Latro, Heldmann (1979) shows that the phrase probably refers to Demetrius’ use of extraordinary and unusual expressions, that is, periphrasis instead of direct expression. For a similar assessment of Demetrius’ style, see Quint. Inst. orat. 1.10.80 (= Demetrius of Phalerum F180 Wehrli). On softness and indirect expressions as attributes typically associated with Asianism, see esp. section 5.3 below. Chiron (2014) 120 points out that Demetrius has collaborated with the dictators of Athens, which Cicero condemns in Brut. 70 Cic. Brut. 37: Delectabat magis Atheniensis quam inflammabat. Ibid.: Non tam armis institutus quam palaestra. Ibid. 38: Suavis videri maluit quam gravis, sed suavitate ea, qua perfunderet animos, non qua perfringeret. Cf. Cicero’s comparison between rhetorical and philosophical prose: the former is considered tough and direct, whereas the latter is ‘soft and academic’ ( mollis et umbratilis ). See Dion. Hal. Dem. 32.1, comparing Plato’s style to ‘bodies that pursue a life of ease in the shade’ ( σκιὰς καὶ ῥᾳστώνας διώκοντα σώματα ). 71 On the chronological issues concerning Atticism, see section 1.4 above. On the problematic treatment of Atticism in modern scholarship, see section 5.2 below. For the relationship between Cicero’s, Dionysius’ and Calvus’ conceptions of Attic style, see esp. sections 5.4–5.6 below.

122

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Atticists’ insisted that oratorical style should imitate the simplicity of Lysias and should therefore be devoid of extravagant ornamentation: accordingly, they attacked Cicero for his lavish use of rhythm and periodic sentence structure (section 5.6.1). In his Orator , the latter offered a sustained, theoretically underpinned response to the Atticists’ minimalistic view of style: according to Cicero, the power of oratory is to a large extent based on style and an orator should therefore gain access to all of its resources. After all, the word eloquentia itself, he argues, highlights ‘style’ ( elocutio ) itself, the third task of the orator: ‘The very name ‘eloquent’ shows that the orator excels because of this one quality, that is, in his use of language, and the other qualities are overshadowed by this. For the all-inclusive word is not discoverer ( inventor ) or arranger ( compositor ) or performer ( actor ), but in Greek he is called ῥήτωρ form the verb ‘to speak’, and in Latin he is said to be eloquens .’72 Cicero’s insistence on the importance of style in his Orator is an integral part of his self-defense against the scorn of the frugal Atticists. The apology of his own stylistic legacy is to a large degree based on the theory of three styles. In fact, he uses the threefold division to attack the self-styled Attic orators in three ways. First, he claims that the Atticists are misguided in their focus on one stylistic register only, as there are actually three separate styles which all deserve to be labeled ‘Attic’. According to Cicero, the ideal orator should, like Demosthenes, be a master of all three styles (section 2.3.2). Secondly, since the Atticists only accept the simple style as genuinely Attic, the Roman author closely examines this style and concludes that their conception of it is misguided. And thirdly, Cicero argues that the best style is the one that is diametrically opposed to the one favored by the Atticists, in spite of which Cicero insists that it is still thoroughly Attic. 73 In a word, the theory of three styles gives Cicero ample opportunity to refute his opponents’ claims about Attic style and to defend his own oratorical legacy. We should note, however, that Cicero does not always use the three-style formula: in his Brutus , which was published in the early months of 46 BC, briefly before his Orator

72 Cic. Orat. 61: Quem hoc uno excellere, id est oratione, cetera in eo latere indicat nomen ipsum. Non enim inventor aut compositor aut actor qui haec complexus est omnia, sed et Graece ab eloquendo ῥήτωρ et Latine eloquens dictus est. 73 Dugan (2001) 406–420 and (2005) 270–288 highlights this point, showing that Cicero uses the three-style doctrine mainly to contrast the plain with the grand, the Atticists’ meager style with his own extravagant style: ‘We find Atticisim and Asianism along the familiar polarities of thin and fat, dry and moist.’ In this section, I will show that the three styles do not merely highlight the contrast between Calvus’ frugal and Cicero’s abundant oratory, but that they provide Cicero with various arguments to refute the claims of his opponents.

123

CHAPTER THREE

(section 1.4), Cicero rather seems to apply a two-style division. Comparing the styles of C. Aurelius Cotta (124–73 BC) and P. Sulpicius Rufus (124/123–88 BC), two famous orators from Cicero’s youth, he notes: ‘Since then there are two distinct types of good oratory—and that is the only kind we are considering—one simple and concise, the other elevated and abundant.’ 74 This is the only explicit reference to stylistic registers in the entire Brutus : its conspicuous omission of the intermediate style has left modern readers astonished. 75 In my view, the twofold doctrine of Brutus and the threefold formulas of De oratore and Orator testify to the flexibility of the discourse on prose style in Late-Republican Rome. In his Brutus , Cicero merely contrasts the styles of two famous orators, but in his Orator , the same author is concerned with the importance of stylistic versatility in the face of the monolithic conception of style of the self-proclaimed Atticists. The malleability of stylistic theory allows him to present only two styles in the former treatise, while he adds a third style in the latter work. In what way does Cicero exploit the three-style formula in his Orator in order to undercut the anti-Ciceronian arguments of the Atticists? First, as I have noted, Cicero insists that the consummate orator can employ all three styles at will. He connects the three styles, as we have seen, to the three tasks of the orator: the simple style is used for proof ( in probando ), the middle style for pleasure ( in delectando ), and the grand style to sway the audience (in flectendo ). 76 To ignore two of the three styles, as Cicero accuses the Atticists of doing, is to

74 Cic. Brut. 201: Quoniam ergo oratorum bonorum, hos enim quaerimus, duo genera sunt, unum attenuate presseque, alterum sublate ampleque dicentium. Cotta’s style is considered typical for the former type, while Sulpicius’ style is thought to be characteristic for the latter type. Both men feature as interlocutors in De or. 75 Hendrickson (1905) 270–271 argues that Brut. and Orat. tap into different traditions concerning the types of style, the Brut. referring to the older view, which goes back to Aristotle, that there are two diametrically opposed styles, while Orat. accepts the more recent addition of a third, intermediate style. Quadlbauer (1958) and Fantham (1979) 450 regard the two-style theory of Brut. as a temporary deviation from Cicero’s subscription to the three-style formula. Narducci (2002) 404, next, tries to reconcile Brut. with Orat. by arguing that Cicero in the former work already propagates the ideal of stylistic versatility. Guérin (2014), lastly, thinks that Cicero’s conception of the stylistic registers is subject to a development in three stages: in De or. , first, the author casually distinguishes between three types of style; in his Brut. , next, he presents a normative doctrine of style which is limited to two types; in his Orat. , lastly, Cicero distinguishes between three styles, so as to illustrate his ideal of stylistic versatility (‘polyvalence stylistique’), replacing the notion of various types of style with the idea of one (or three-in-one) perfect style. I contend that the differences between De or. , Brut. , Orat. can be more elegantly and less problematically ascribed to the flexibility of the stylistic discourse in the first century BC. 76 Cic. Orat. 69. See section 3.2.1, esp. table 3, above for the ways in which Cicero aligns his articulations of the three-style formula in Orat. to his triple doctrine of oratorical tasks. Cf. also Guérin (2014) 168–171.

124

THE PLAIN , THE GRAND AND THE IN-BETWEEN neglect two essential tasks of the orator. Hence, the Atticists can provide proof for their arguments, but they are unable to entertain or sway their listeners. The perfect orator, Cicero claims, ‘will decide what is needed at any point, and will be able to speak in any way which the case requires’. 77 Appropriateness ( τὸ πρέπον , or, as Cicero translates it, decorum ) is key and hinges on several aspects—the subject under discussion, the character of both speaker and audience, the section of the speech to which the words belong (table 3).78 All these aspects should be considered in determining which type of style is required. In Cicero’s view, the Atticists cannot speak appropriately on every occasion, for they lack the stylistic flexibility to speak in any way which the case demands. Secondly, Cicero dwells extensively on the plain style, which, after all, is the only true Attic style according to Calvus and his followers. Cicero’s discussion of the characteristic features of the simple style aims to show that the Atticists misunderstand the nature of this register. To begin with, Cicero distinguishes between two types of ‘plain’ orators: some are ‘astute’ ( callidi ) but ‘unpolished’ ( impoliti ) and ‘intentionally resembling the speech of untrained and unskillful speakers’ ( consulto rudium similes et imperitorum ); others, conversely, are ‘more balanced’ ( concinniores ), ‘elegant’ ( faceti ), even ‘flourishing’ (florentes ) and ‘to a slight degree ornate’ ( leviter ornati ). 79 The self-proclaimed Atticists belong to the first group: according to Cicero their rough and unpolished speech is equivalent

77 Cic. Orat. 69: Nam et iudicabit quid cuique opus sit et poterit quocumque modo postulabit causa dicere . Cf. the definition of the ideal orator provided by Crassus in De or. 1.64: ‘That man, in my opinion, will be an orator, worthy of so great a name, who, whatever subject comes before him, and requires rhetorical elucidation, can speak on it judiciously, in set form, elegantly, and from memory, and with certain dignity of action’ ( is orator erit mea sententia hoc tam gravi dignus nomine, qui, quaecumque res inciderit, quae sit dictione explicanda, prudenter et composite et ornate et memoriter dicet cum quadam actionis etiam dignitate ). Transl. May and Wisse (2001). Here, the difference in scope between De or. and Orat. comes to the fore: whereas the former work looks for perfection in all five canons of eloquence, the latter work focuses on style. 78 For Cicero’s views of appropriateness, see Or. 70–74. Pohlenz (1933) explores the evolution of the meaning of τὸ πρέπον , as a crucial concept in Greek thought. The locus classicus for appropriateness in stylistic theory is Arist. Rh. 3.2: see section 3.1 n. 8 above. Cicu (2000) focuses on the relationship between the Ciceronian notion of decorum and the Greek concept of τὸ πρέπον : cf. section 3.2.1 n. 27 above. It appears that Cicero’s interpretation of literary appropriateness is related to his view of moral appropriateness: Cic. Orat. 70 professes that the same law holds good ‘in life as well as in speech’ ( ut enim in vita, sic in oratione ). Decorum is not only Cicero’s fourth out of four cardinal stylistic virtues, but also his fourth out of four cardinal virtues in life itself: see Off. 1.93–151. 79 Cic. Orat. 20: In eodemque genere alii callidi, sed impoliti et consulto rudium similes et imperitorum, alii in eadem ieiunitate concinniores, idem faceti, florentes etiam et leviter ornati .

125

CHAPTER THREE to an ‘incapacity of coherent speech’ (infantia ), but the trick is to appear to speak in the language of the masses without actually abasing oneself to the inadequacy of the speechless (infantes ). 80 Therefore, even the orator of the plain style should not wholly eschew stylistic ornamentation: ‘Although it is not full-blooded ( non plurimi sanguinis ), it should nevertheless have some of the sap of life ( sucus ) so that, though it lacks supreme power ( maximae vires ), it may still be, so to speak, of sound health ( integra valetudine ).’ 81 The Rhetorica ad Herennium already describes the faulty style that is akin to the simple style as ‘bloodless’ ( exsanguis ), ‘dry’ ( aridus ) and ‘meager’ ( exilis ); Cicero ascribes a similar fault to the Atticists. 82 It is clear, then, that Cicero prefers a simple style that is not entirely devoid of stylistic flourish. What ornamentations are allowed in Cicero’s conception of the plain type of oratory? He admits that some devices are taboo, such as rhythm, ‘cementing the words together’ ( verba verbis coagmentare ) by avoidance of hiatus, and periodic structure. 83 Yet, the orator is allowed to make use of the ‘orator’s toolbox’ ( oratoria supellectilis ), provided that he do so moderately and inconspicuously. 84 Cicero compares the simple style to a woman, who is prettier when she remains unadorned. In dressing up this lady, all noticeable ornaments, such as jewelry and cosmetics, should be removed, so that elegance and neatness (elegantia et munditia ) remain: the plain-style orator should not use ‘pearls’ ( margaritae ), ‘curling-irons’ ( calamistri ), or ‘artificial white and rouge’ ( medicamenta candoris et ruboris ). 85 What, then, is left in the rhetorical toolbox for the orators of the simple style to

80 Cic. Orat. 236. Cf. ibid. 76, explaining that an audience of ‘people who do not possess the faculty of natural speech’ (infantes ) tend to think that an oration in the simple style is ‘easy to imitate’ ( imitabilis ): yet, ‘nothing is less true, when one attempts it’ ( nihil est experienti minus ). Similarly, Dionysius praises Lysias, despite the simplicity of his style for his artistry: cf. n. 87 below. 81 Cic. Orat. 76: Etsi enim non plurimi sanguinis est, habeat tamen sucum aliquem oportet, ut, etiamsi illis maximis viribus careat, sit, ut ita dicam, integra valetudine . 82 Rhet. Her. 4.16. Demetr. Eloc . 236–239 presents the so-called ‘arid’ ( ξηρός ) style as a neighboring vice to the simple style, while Gell. NA 6.14.5 calls the speakers who unsuccessfully aim at simplicity ‘filthy and barren’ (squalentes et ieiuni ). See section 3.2.1 n. 26 for the ancient sources on faulty types of style. 83 The theory of word arrangement is the second major topic of Orat. , in addition to the theory of three styles. For Cicero’s ambivalent approach to the topic of hiatus, see Orat. 77–78, 150–152 with section 4.5.1 below. 84 Cic. Orat. 80. 85 Cic. Orat. 78–79. Leidl (2003) discusses Cicero’s representation of rhetoric as a casta virgo (e.g. Brut. 353, Orat. 64) and links it to Dionysius’ images of the chaste muse representing Attic oratory and the insensate harlot representing Asianic rhetoric at Ant. orat. 1.5–7. On Dionysius’ allegory see esp. De Jonge (2014a). Cf. section 5.4 below on the personifications of Attic style in Cicero and Dionysius: we will see that the proposed imagery makes the various conceptions of Atticism in Rome conveniently tangible.

126

THE PLAIN , THE GRAND AND THE IN-BETWEEN use? 86 They are allowed, for one, to use metaphors, as long as they are not far-fetched and only if they are used to make the meaning clear: they must not be meant solely for entertainment. Furthermore, the simple orator may add figures of speech to his discourse, provided that they do not result in periodic sentences or involve uncommon words and harsh imagery, and he may also use figures of thought, if they are not ‘overly glaring’ (vehementer illustria ). 87 Lastly, Cicero requires the plain orator to sprinkle humor and wit ( facetiae et dicacitas ) over his speech: he considers these features hallmarks of Atticism ( maxime Atticum ), although he has never seen any of ‘these modern Atticists’ ( isti novi Attici ), that is, Calvus and his followers, use them correctly. 88 The third way in which Cicero redefines Attic oratory is concerned with the grand style. As he argues in his Orator , this grand style is the most powerful style in public oratory: ‘For it is the one thing that avails the most in winning cases.’ 89 Between his interest in the grand and the simple style Cicero virtually glosses over the middle style, claiming, as we have seen, that it does not originally come from the Forum, but from the schools of the philosophers and the sophists, and that it is better suited to epideictic oratory than to the action of the courts and the assemblies (section 3.2.2). The middle style holds a rather ambiguous mean between grandeur and simplicity: ‘Between these two there is a mean and I may say tempered style, which uses neither the intellectual appeal of the latter class nor the thunderous force of the former; akin to both, excelling in neither, sharing in both, or, to tell the truth,

86 Cic. Orat. 81–90 describes the rhetorical tools that are permitted for the simple orator, discussing metaphors (81–82), figures of speech and thought (83–85) and humor (86–90) respectively. 87 The gist of the matter seems to be that the orator may use stylistic ornamentation, as long as it is not obvious to his audience that he does so. Cf. De Jonge (2008) 224–226 on Dionysius’ views of natural style, esp. with respect to Lysias: ‘He does not mean to say that Lysias’ composed his speeches instinctively, nor that he did not make use of artistic techniques. In fact, Lysias’ speeches are supposed to be the product of an art ( τέχνη ) that imitates nature ( φύσις ).’ See, e.g., Dion. Hal. Lys. 8.6: Lysias’ arrangement of words ‘is more carefully composed than any work of art. For this artlessness is itself the product of art: the relaxed structure is really under control, and it is in the very illusion of not having been composed with masterly skill that the mastery lies’ (ἔστι δὲ παντὸς μᾶλλον ἔργου τεχνικοῦ κατεσκευασμένος . Πεποίηται γὰρ αὐτῷ τοῦτο τὸ ἀποίητον καὶ δέδεται τὸ λελυμένον καὶ ἐν αὐτῷ τῷ μὴ δοκεῖν δεινῶς κατεσκευάσθαι τὸ δεινὸν ἔχει ). Cf. Viidebaum (2019) 110 –111, who notes that Dionysius’ emphasis on the deceptive quality of Lysias’ style resembles Phaedrus’ judgment of the orator’s style in Pl. Phdr. 227c5, 228a2. 88 Cicero uses the distinction between facetiae and dicacitas as an apology for Demosthenes’ perceived deficiency in the field of humor: see section 2.3 n. 35 above. Wisse (1995) 75–76 adduces the phrase ‘these modern Atticists’ ( isti novi Attici ) as proof for ‘a date of ca. 60 BCE for the beginning of Atticism as a whole’. 89 Cic. Orat. 69: Nam id unum ex omnibus ad obtinendas causas potest plurimum.

127

CHAPTER THREE sharing in neither.’ 90 As a result of this vague description, the middle style has been called a ‘colorless half-way house’. 91 Yet, it adds substance to Cicero’s argument about the versatility of Attic style, and, by connecting this vague mean to his doctrine of the orator’s tasks, he can use it as yet another stick to beat his opponents with. To return to the grand style, we have seen that this elevated type is typically characterized by ‘force’ (vis ). Cicero argues that this style has ‘the greatest power’ ( maxima vis ) of all three styles and even that it sums up ‘all the orator’s power’ ( vis omnis oratoris ). 92 According to the rhetorician, Demosthenes receives the most lavish praise, when he expends the full power of his eloquence.93 There is, however, an important caveat: an orator must only use the grand style at appropriate occasions and even then he must temper its abundance with the other two styles. If he launches into a fiery, vehement speech, without preparing the minds of his listeners, ‘he seems to be a raving madman (furere ) among the sane, like a drunken reveler (bacchari ) in the midst of sober men.’ 94 An orator, who uses one of the other two styles, ‘is far from standing on slippery ground, and, once he gets a foothold, he will never fall’.95 Cicero’s emphasis on the dangers of grandiloquence accords well with ancient criticism on sublime literature, which stresses the perils of elevation. The author of On the Sublime , for example, submits that ‘it may be inevitable that humble, mediocre natures, because they never run any risks and never aid at the heights, should remain to a large extent safe from error, while in great natures their very greatness spells danger’.96

90 Cic. Orat. 21: Est autem quidam interiectus inter hos medius et quasi temperatus nec acumine posteriorum nec fulmine utens superiorum, vicinus amborum, in neutro excellens, utriusque particeps vel utriusque, si verum quaerimus, potius expers. 91 Winterbottom (1989) 127. Cf. Leeman (1963) 39 on the discussion of the middle style in Rhet. Her. : it ‘is mainly characterized by what it is not. It does not have the pathos of the gravis figura , nor the naked simplicity of the figura adtenuata . What remains is a pleasant and relaxed loftiness’. On the middle style in Cic. and Rhet. Her. , see esp. section 3.2.2 above. 92 Cic. Orat. 69, 97. 93 For the ancient tradition of associating Demosthenes particularly with power and forcefulness ( δεινότης , vis ), see section 2.3.1, esp. n. 47, and section 2.3.2, esp. n. 59, above. 94 Cic. Orat. 99: Furere apud sanos et quasi inter sobrios bacchari vinulentus videtur. Long. Subl. 3.2, 16.4 and 32.7 also uses the follower of Bacchus as an image for the author of sublime literature. See also De Jonge (2012a), who notices a remarkable convergence in Dionysius’ and Longinus’ use of religious language to describe elevation or sublimity. 95 Cic. Orat. 98: Mimimeque in lubrico versabitur et, si semel constiterit, nunquam cadet. 96 Long. Subl. 33.2: Μήποτε δὲ τοῦτο καὶ ἀναγκαῖον ᾖ, τὸ τὰς μὲν ταπεινὰς καὶ μέσας φύσεις διὰ τὸ μηδαμῇ παρακινδυνεύειν μηδὲ ἐφίεσθαι τῶν ἄκρων ἀναμαρτήτους ὡς ἐπὶ τὸ πολὺ καὶ ἀσφαλεστέρας διαμένειν , τὰ δὲ

128

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Cicero agrees with the later Greek authors Dionysius and Longinus that the grand style has the power to carry away the minds of the audience and to move them in any possible direction, provided that it is used appropriately. If the stakes are highest, Cicero argues, the grand style is required. As I have already noted, he associates the grand style with Pericles or Demosthenes, who had used their eloquence for the noblest and grandest of causes, viz., the defense of Greek freedom: we will see that Cicero imagines himself a Roman-era Pericles and Demosthenes in the face of Caesar’s dictatorship (section 5.5.1). In his essay De Optimo Genere Oratorum , moreover, he remembers the circumstances of his speech in defense of Milo (52 BC), ‘when the army was stationed in the Forum and in all the temples round about’: he censures anyone who thinks that in such a dangerous situation, a subdued, Lysianic style would suffice. In other words, neither Calvus nor any of his fellow-Atticists would have been able to pull such a life-and-death speech off. 97 Ironically, Cicero’s effort ultimately did not secure an acquittal for Milo: it is the rewritten version of the speech that has come to be celebrated as one of the most accomplished orations in Cicero’s rhetorical oeuvre. 98 We may conclude that the theory of three styles provides Cicero with several powerful strategies against the self-styled Atticists. He uses it to show the variety of Attic oratory, to correct the Atticists’ interpretation of the simple style, and finally to point out where the true power of eloquence can be found—on the dangerous cliffs of sublime oratory.

3.5 Dionysius and the Aesthetics of the Middle Naturally, Dionysius does not use the three-style formula necessarily for the same purposes as Cicero. As was the case with Cicero, however, we will see that Dionysius’ conception of the

μεγάλα ἐπισφαλῆ δι ’ αὐτὸ γίνεσθαι τὸ μέγεθος . See Porter (2016) 198–199 for the connections between the sublime and danger in ancient literary criticism: the risky nature of elevated style is already discussed in Pl. Resp. 497d9–10. Dion. Hal. Dem. 2.6 considers ‘risk’ ( τὸ τολμηρόν ) a feature of elevated diction, and ‘safety’ (ἀσφάλεια ) a feature of the plain type. The mistakes most frequently associated with grandeur are ‘frigidity’ (ψυχρότης ) and ‘bombast’ ( τὸ ὄγκος ): see esp. Gutzwiller (1969). Cf. section 2.4.1 n. 110 above. 97 Cic. Opt. gen. 10: Exercitu in foro et in omnibus templis, quae circum forum sunt, collocato . See section 2.3.2 n. 61 for a fuller discussion of this passage. La Bua (2010) 36–37 argues that Cicero refers to Mil. in Opt. gen. , because he could not leave unanswered the idea that he was unable to deliver a good speech in a tough situation. 98 For this lavish praise, see Asc. Mil. 42c: ‘He has written the text that we can read so perfectly that it can rightly be considered his best’ ( scripsit vero hanc quam legimus ita perfecte ut iure prima haberi possit ). Dio Cass. 40.54.3 relates that, when Milo read Cicero’s rewritten speech during his exile, he bitterly remarked ‘that, if Cicero had spoken in this manner, he would not now be enjoying the delicious red mullets of Massilia’ ( οὐ τοιαύτας ἐν τῇ Μασσαλίᾳ τρίγλας ἐσθίειν , εἴπερ τι τοιοῦτον ἀπελελόγητο ).

129

CHAPTER THREE three stylistic registers is to a large extent accommodated to suit the specific purposes of his essays. Significantly, he expounds the three styles in only two extant works, On Demosthenes and On the Arrangement of Words , both attributed to his so-called middle period. In the former work the threefold formula is omnipresent: Dionysius discusses both the three types of diction and the three types of arrangement.99 We find no mention of these divisions in the remainder of his extant oeuvre, bar one quote from On Demosthenes in the Letter to Pompeius .100 He does not refer to the three styles, for instance, in the surviving fragments of On Imitation nor in his other treatises on individual authors, such as On Lysias , On Isocrates and On Thucydides . Yet, he presents Lysias, Isocrates and Thucydides in On Demosthenes and On the Arrangement of Words , as we have seen, as outstanding exponents of the three styles (section 3.3). Why, then, does he so suddenly and so intensely turn to the three styles in the latter two treatises, while he does not show any interest in the topic in his other works? To answer this question we must first briefly address the works in which Dionysius does not concern himself with the threefold division: in these works the critic instead adopts a twofold scheme. We have already seen that Cicero resorts to such a binary view, when he discusses the styles of Sulpicius and Cotta in his Brutus (section 3.4): likewise, Dionysius only recognizes two types of style, as long as he is discussing the stylistic character of only one or two authors. Concerning diction, for instance, he often contrasts plain to elevated language, omitting the intermediate type. Both his earlier works, such as On Lysias , and his later works, such as On Thucydides , exemplify this binary approach: Dionysius adduces the contrast between plain and elevated diction, for instance, to throw into relief Lysias’ simplicity and Thucydides’ loftiness. 101 In On Isocrates , moreover, Dionysius does not classify Isocrates’ style as mixed, as he does in On Demosthenes , but he rather sets up the orator in a binary scheme against Lysias: he presents Lysias, unsurprisingly, as the more down-to-earth orator, while Isocrates is described as ‘more elevated, more impressive and

99 For the types of diction, see Dem. 1–34; for the types of arrangement, see Comp. 21–24 and Dem. 35–52. These two treatises were composed in roughly the same period: for the relative chronology of Dionysius’ critical oeuvre, see section 1.5 above. 100 Pomp. 2 quotes Dem. 5–7 (on Plato’s mixed style of diction) almost verbatim. The fact that theories of three styles appear only in Dem. , Comp. and Pomp. is a strong argument in favor of studying these works as a separate group within Dionysius’ oeuvre: cf. Bonner (1939) 32, De Jonge (2008) 19–20. 101 Cf. Hendrickson (1904) 143: for Dionysius, ‘there are but two fundamental styles, the simple and the grand.’ Dion. Hal. Lys. 8.3, for instance, contrasts ‘clear, standard, ordinary speech which is thoroughly familiar to everyone’ ( ἡ σαφὴς καὶ κυρία καὶ κοινὴ καὶ πᾶσιν ἀνθρώποις συνηθεστάτη ) to ‘pompous, outlandish and contrived language’ ( ὁ ὄγκος καὶ τὸ ξένον καὶ τὸ ἐξ ἐπιτηδεύσεως ). Cf., e.g., ibid. 13.2–5, Thuc. 33.2.

130

THE PLAIN , THE GRAND AND THE IN-BETWEEN more dignified’ ( ὑψηλότερος , μεγαλοπρεπέστερος , ἀξιωματικός ). 102 Apparently, the middle type of diction is not yet on Dionysius’ mind. As for the arrangement of words, the situation is similar: the critic has no attention for an intermediate harmony in the majority of his oeuvre, but he is only interested in binary divisions, pitting Isocrates’ smooth, artificial harmony indiscriminately against Thucydides’ rough type as well as against Lysias’ simple style. 103 Thus, the feature that sets apart On Demosthenes and On Composition from the rest of Dionysius’ critical works is the introduction of a mixed, intermediate style between the two opposite extremes, that is, either between simple and elevated diction, or between smooth and rough arrangement. In both topics of style, Dionysius declares the middle type to be the most powerful style: he consistently classifies his favorite authors in this category. On the subject of word selection, his favorite is Demosthenes, who ‘perfected’ ( ἐτελείωσεν ) the middle style after he had inherited it ‘in an imperfect form’ ( ἀτελές) from Plato and Isocrates.104 On the subject of word arrangement, next, his absolute favorite is Homer, ‘the summit on which everyone’s gaze should be fixed’ ( κορυφὴ ἁπάντων καὶ σκόπος ), while in prose he considers Herodotus, Plato and Demosthenes second to none.105 The middle style provides Dionysius with a framework within which he can describe the perfect style. It is not a coincidence that we find it only in On Demosthenes and On the Arrangement of Words : it is in these two works that Dionysius sets out to identify his most beloved literary heroes and the stylistic features that set them apart. What makes the middle style such a convenient vehicle for Dionysius for articulating his ideas about stylistic excellence? I have already noted that Dionysius actually interprets the concept of the middle in two different ways (section 3.2.2): he conceives it not only as an intermediate stage between two extremes ( μέσος , μεταξὺ τῶν ἄκρων ), but also as a kind of mixture or combination of these very extremes ( μῖγμα , κρᾶσις , σύνθεσις ). 106 Dionysius

102 Dion. Hal. Isoc. 3.5. 103 For Isocrates’ arrangement, see Dion. Hal. Isoc. 2.4–7, 12.3–13. It is not illogical that Dionysius contrasts Isocrates to both Thucydides and Lysias: as we have seen in section 3.3, n. 62 above, Lysias’ uncontrived, natural style of word arrangement displays several features that Dionysius associates with rough harmony. 104 Dion. Hal. Dem. 14.1. 105 See Dion. Hal. Comp. 24.4 for the praise of Homer’s arrangement, and Dem. 41.2–3, 43.1; Comp. 10.5, 19.12 and 25.29–44 for the praise of the arrangement of Herodotus, Plato and Demosthenes. Among the latter three prose authors, Demosthenes is Dionysius’ favorite overall author (cf. section 2.3.3 above), but on the specific topic of word arrangement, Dionysius declares Plato to be at the very least Demosthenes’ equal ( Comp. 18.13). 106 These two, seemingly contradictory, interpretations of the middle style are subject of some discussion. Hendrickson (1904) 143, for instance, argues that the two conceptions of the middle in Dionysius’ works are

131

CHAPTER THREE himself seems to be rather indifferent about the potential incompatibility between these two interpretations: ‘I have no idea how to describe the way in which the third style is produced —“my mind is too divided to utter the truth”. I cannot say whether it is formed by removing the two extremes ( κατὰ στέρησιν ) or by combining them ( κατὰ μῖξιν ), for it is not easy to find a clear solution to problem.’ 107 The two ideas often appear side by side: the middle type of word selection is said to be ‘intermediary and well-blended’ ( μέση καὶ εὔκρατος ) and the middle type of word arrangement is described as ‘intermediary and mixed’ ( μέση καὶ μικτή ). 108 In the present section, we will see that Dionysius is not at all worried about the consistency of his teachings: he simply puts the discrepant interpretations to use as complementary strategies to present his stylistic program. 109 First of all, by situating the middle style simply between the other two styles Dionysius can add the authority of Aristotle and Theophrastus to his argument. This indebtedness of Dionysius to the Peripatetic school has often been remarked upon. 110 incompatible, the idea of a mixture bearing ‘only a crude external resemblance’ to the Peripatetic concept of a mean between the extremes. Likewise, Stroux (1907) 111 n. 3 thinks that ‘the former interpretation in fact contradicts the latter’ ( revera altera impugnat alteri ), adding that Dionysius’ mixture can be linked to the idea of a proper mean through his insistence on good measure. Bonner (1938) 263–264 argues that the idea of a mixed style follows from a ‘quite understandable development’ from Aristotle’s and Theophrastus’ argument for a mean between the extremes. Goudriaan (1989) 504–510, lastly, attempts to reconcile Dionysius’ and Aristotle’s views on style by quoting Arist. Poet. 22.1458a31–34 and Rh. 3.12.6, which both refer to stylistic mixtures. 107 Dion. Hal. Comp. 21.4–5: Ἣν ὅπως ποτὲ γίνεσθαι φαίην ἄν, ἔγωγε ἀπορῶ καὶ ‘δίχα μοι νόος ἀτρέκειαν εἰπεῖν’, εἴτε κατὰ στέρησιν τῶν ἄκρων ἑκατέρας εἴτε κατὰ μῖξιν· οὐ γὰρ ῥᾴδιον εἰκάσαι τὸ σαφές. The pronoun ἥν refers to the middle type of arrangement ( ἡ τρίτη κοινὴ διαφορά ). The quote is Pind. fr. 213 Schroeder. 108 Dion. Hal. Dem. 3.5, 42.2. See also ibid. 36.6, 43.10 and Comp. 24.1. Cf. Cic. Orat. 21, which describes the intermediate style as ‘sharing in both, or, to tell the truth, sharing in neither’ ( utriusque particepts vel utriusque, si verum quaerimus, potius expers ): Cicero, however, does not explore the notion of mixing any further: while Dionysius defines the middle style always in relation to the other two styles, Roman authors tend to assign to the middle style its own particular character and function, that is, to provide pleasure and entertainment. Quint. Inst. orat. 12.10.58, as we have seen, gives ‘flowery’ ( floridum ) as an alternative name for the middle style. 109 Porter (2016) 221 shows that Dionysius never meant his categories to be taken ‘in a hard-and-fast way’ and that ‘the various surface inconsistencies in Dionysius’ theory and evaluative practice are best explained by the provisional nature of his schemas’. It is my contention that the ‘surface inconsistencies’ are conveniently put to use by Dionysius as sources for various ad hoc arguments in favor of his own critical program. 110 For the debate about Aristotle’s successor Theophrastus as the possible inventor of the theory of three styles, see section 3.1 n. 9 above. For the relationship between Dionysius’ theory of three styles and the Aristotelian notion of the mean, see the present section n. 106 above. To these discussions add Kroll (1907) 91–101, Pohl (1968) 94, both suggesting Peripatetic origins for Dionysius’ views on musical word arrangement, and Paximadi

132

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Dionysius himself explicitly links his views to Aristotle, when he discusses the intermediate type of word arrangement: he argues that this style deserves the first price, ‘since it represents a sort of mean ( μεσότης τις ), and virtue in life, conduct and the arts is a mean in the view of Aristotle and the other philosophers of his school.’ 111 The old master had submitted that in any aspect of life and the arts there exist ‘excess’ ( ὑπεροχή), ‘deficiency’ ( ἔλλειψις ) and finally a ‘mean’ ( μέσον ), which is where excellence (ἀρετή ) can be found. In this vein, Aristotle prescribes in his Rhetoric , as we have seen, that stylistic excellence ( ἀρετὴ λέξεως ) means that style should be ‘neither flat (ταπεινή ) nor above (ὑπέρ ) the dignity of the subject, but appropriate ( πρέπουσα ).’ 112 Aristotle’s successor in the Lyceum, Theophrastus, seems to have thought along the same lines about style, as far as we can tell from the scattered remains of his works.113 Like the old masters of the Lyceum, Dionysius stresses the importance of the appropriate mean: he proclaims that appropriateness ( τὸ πρέπον ) is ‘the most important of all literary qualities’ ( πασῶν ἐν λόγοις ἀρετῶν ἡ κυριωτάτη ). 114 He considers the stylistic extreme types of diction and arrangement ‘imperfect’ ( ἀτελεῖς) in this respect. 115 He describes

(1989) 223–225, who connects Dionysius’ grammatical theories to Peripatetic sources. In my view, as I will explain below, Dionysius’ subscriptions to the teachings of a certain philosophical school (Peripatos, Stoa, etc.) are quite opportunistic: he can unabashedly abandon the views of his predecessors, if that suits his purposes. 111 Dion. Hal. Comp. 24.2: Ἐπειδὴ μεσότης μέν τίς ἐστι (μεσότης δὲ ἡ ἀρετὴ καὶ βίων καὶ ἔργων καὶ τεχνῶν, ὡς Ἀριστοτέλει δοκεῖ τε καὶ τοῖς ἄλλοις ὅσοι κατ ’ ἐκείνην τὴν αἵρεσιν φιλοσοφοῦσιν . 112 A good starting point for studying Aristotle’s conception of μεσότης is Moreau (1962) 207, who prints and discusses such key passages as Arist. Eth. Nic. 2.6.4–2.7.1, Pol. 3.8.5–7, and Eth. Eud. 2.2–2.3. For a discussion of Arist. Rh. 3.2, see section 3.1 n. 8 above. 113 Two Theophrastean fragments specifically pertain to the middle as a stylistic concept. Dion. Hal. Dem. 3.1 (= Theophr. fr. 685 Fortenbaugh) shows that Theophrastus considered Thrasymachus’ prose a ‘source for the mean’ (πηγή τις τῆς μεσότητος). Cf. section 3.1 n. 10 above and the secondary literature cited there. Demetr. Eloc. 114 (= Theophr. fr. 686 Fortenbaugh) reports that Theophrastus describes frigidity ( τὸ ψυχρόν ) as ‘that which exceeds the proper form of expression’ ( τὸ ὑπερβάλλον τὴν οἰκείαν ἀπαγγελίαν ). Innes (1985) 260–262 and Fortenbaugh (2005) 273–281 offer thoughtful discussions of these passages. 114 Dion. Hal. Pomp. 3.20. Cf. Comp. 12.20, which declares appropriateness is also the ‘most powerful’ ( πάντων κράτιστον ) means of composition. See also section 3.2.1 n. 8 and section 3.4 n. 78 above for Cicero’s view on the importance of appropriateness ( decorum ) and its relationship to Greek views. 115 Dion. Hal. Dem. 2.8 states that Lysias and Thucydides ‘are both brilliant in their own specialties, but also— and in this respect they are similar to each other—imperfect’ ( δεινοὶ μὲν ἐν τοῖς αὑτῶν ἔργοις ἀμφότεροι , καθ ’ ὃ δὲ ἴσοι ἀλλήλων ἦσαν , ἀτελεῖς). The translation, which follows Bonner (1938) 252 and Aujac (1988) 50, is mine. Note that Usher (1974) 247, agreeing with Hendrickson (1904) 142, translates the last part of the passage as ‘imperfect in respect of those qualities which they possess in common’ ( καθ ’ ὃ δὲ ἴσοι ἀλλήλων ἦσαν ,

133

CHAPTER THREE their faults, in accordance with Aristotle, as excesses and deficiencies: Thucydides is said to use stylistic ornamentation ‘without moderation’ ( ἀταμιεύτως ) and ‘excessively’ (κατακόρως ), while Lysias is sometimes censured for being too ‘faint’ (ἀμυδρός ) and ‘feeble’ (ἀσθενής ). Plato and Isocrates, conversely, are classified in the superior middle category, although in Dionysius’ view their prose is ‘imperfect’ ( ἀτελές ) as well. 116 He portrays their mistakes, again, as failures in observing the appropriate mean. He takes issue, for instance, with Plato’s elevated diction: in this register the philosopher is said to display ‘an immoderate eagerness’ ( ἄμετρος ὁρμή ). Isocrates, too, is criticized for not always striking the right note: he sometimes ‘neglects moderation’ ( ὀλιγωρεῖ τοῦ μετρίου ) and ‘loses sight of what is appropriate’ (ἀπολείπεται τοῦ πρέποντος ). 117 Demosthenes, conversely, is celebrated as the consummate orator, precisely because of ‘appropriateness, which touches the stars’ ( τὸ πρέπον ὃ τῶν ἄστρων ψαύει ) in his speeches. 118 Thus, Dionysius turns to Peripatetic vocabulary to give voice to his preference for the middle type of style. We should note, however, that Dionysius is not exclusively committed to Aristotle’s school: he can and does transgress against its teachings, if that suits his purposes. Therefore, Dionysius should not be classified as a Peripatetic, merely because he shares some views with the scholars of the Lyceum. Unlike them, he construes the middle style, as we have seen, as a ‘mixture’ ( μῖγμα ) of the other two styles. 119 In On the Arrangement of Words ,

ἀτελεῖς). Yet, Dionysius does not point out the qualities which Lysias and Thucydides are supposed to have in common: his whole point is to show their incompatibility as stylistic opposites and to point out the extravagance of their extreme styles. 116 For Thucydides, see Dion. Hal. Thuc. 51.3, cf. also Dionysius’ criticism of Gorgias’ elevated style in Is. 19.2, Dem. 4.4 and Thuc. 24.9. For Lysias, see Dem. 13.8. For Plato and Isocrates, see Dem. 14.1. 117 See Dion. Hal. Dem. 5.4, 18.3 and 18.7. Dionysius praises passages from Plato, where the author selects common, everyday words: the critic only censures passages which draw on elevated diction. Concerning Isocrates, however, Dionysius identifies mistakes both in plain and in elevated passage: when Isocrates uses unusual words, the style becomes ‘flowery and showy’ ( ἀνθηρὰ καὶ θεατρική ), while he overzealously strives for clarity in his simple passages, thus not attaining the essential virtue of ‘conciseness’ ( συντομία ). 118 Dion. Hal. Dem. 34.5. The cosmological metaphor can be linked to the ancient discourse on sublime literature: Porter (2016) 537–617 considers such imagery part of the tradition, heavily influenced by Plato, which equates sublimity with ‘the ethereal, otherworldly and immaterial’. In Augustan Rome, Hor. Od. 1.1.35–36 has a similar image: ‘But if you rank me among the lyrical poets, I shall soar aloft and strike the stars with my head’ (quod si me lyricis vatibus inseres, sublimi feriam sidera vertice ). 119 See e.g. Dion. Hal. Dem. 3.1, 5.1, 15.7, 41.1. The words μείγνυμι , μῖγμα , μῖξις and cognates refer to mixtures in a general sense: cf. LSJ s. μείγνυμι . See also n. 124 below on the words κρᾶσις and σύνθεσις , two subtypes of the μιγ -cognates.

134

THE PLAIN , THE GRAND AND THE IN-BETWEEN he also calls the mixed type of arrangement ‘common’ ( κοινός ), expressing the idea that it shares in both the smooth and the rough. 120 Consequently, he does not seem to consider the extreme styles faulty per se: despite their inherent ‘excesses’ ( ὑπερβολαί ) each style has something ‘useful’ ( χρήσιμον ) to offer to the intermediate style. 121 In addition, Dionysius submits that this mixed style has ‘no form peculiar to itself’: it depends entirely on contributions from the other two styles. 122 Thus, Dionysius cherry-picks several ideas from Aristotle and Theophrastus, without committing himself unequivocally to them: the old masters would probably not have approved of their epigone’s interpretation of the middle style. 123 Rather confusingly the critic conveys two divergent ideas as to how this mixture should be formed: he interprets the mixed style occasionally as a ‘blend’ (κρᾶσις ), in which the two extreme styles are merged together to form a homogeneous mixture (like water and wine), but at other times he suggests that the mixture is a ‘combination’ (σύνθεσις ), whose constituents are joined next to each other and remain recognizable as separate styles (like water and oil). 124 How can Dionysius work with these seemingly contradictory notions? 125 In

120 Dion. Hal. Comp. 21.4, 24.1. Aujac and Lebel (1981) 219–220 suggest that Dionysius uses the term ‘common’ ( κοινός ) for the mixed type of arrangement to avoid confusion with the mixed type of diction that he expounds in the first part of Dem. Yet, he is not worried about such confusion in the second part of Dem. , when he never refers to the mixed type of arrangement as ‘common’. 121 See e.g. Dion. Hal. Dem. 3.3, 8.2, 15.7, 41.1; Comp. 21.4, 24.1. Dionysius describes the mixed style as the product of a process of ‘selection’ ( ἐκλογή ) of the ‘most powerful’ ( κράτιστα ) and ‘most useful’ ( χρησιμώτατα ) elements from the extreme styles, at the same time avoiding the excesses. 122 Dion. Hal. Comp. 24.1: σχῆμα ἴδιον οὐδέν . Dem. 41.1: οὐδεὶς χαρακτὴρ ἴδιος . 123 On the incompatibility of Dionysius’ views on the mixed style with the Peripatetic concept of the mean, see the literature cited in n. 106 above. Incidentally, Dionysius’ opportunistic exploitation of various theoretical doctrines, without committing to any philosophical school wholesale, can also be seen in his use of Stoic ideas: see e.g. De Jonge (2008) 240–279 on Stoic and Peripatetic influences on Dionysius’ views of natural word order: it is impossible to ‘draw any conclusions about his alleged philosophy of language’. Indeed, we cannot assign Dionysius to either Peripatos or Stoa on the basis of his own critical essays. 124 For κρᾶσις as a subtype of mixture (for which the μιγ -cognates are used, cf. n. 119 above), see Arist. Top. 4.122b26, Stoic. 2.153. See Dion. Hal. Comp. 2.1 for ‘combination’ ( σύνθεσις ) as a process of ‘placing next to each other’ ( ποιά τις θέσις παρ ’ ἄλληλα ). 125 Like the apparent inconsistency that I discussed in n. 106 above, this issue has also been discussed by several scholars, some interpreting Dionysius’ mixture as a homogeneous blend, others as a mechanical mixture. Pohl (1968) 59–68 and Van Wyk Cronjé (1968) 25 opt for the second view: Van Wyk Cronjé considers Dionysius’ mixed style to be the application of the extreme types, ‘however not simultaneously as mixed might suggest, but in alternating application of either one of the two extremes according to the prerequisite of appropriateness’. As

135

CHAPTER THREE fact, Dionysius needs both interpretations, because one concept works better in word selection and the other in word arrangement. Unfortunately, Dionysius does not express this view unequivocally, as he uses the vocabulary of blending and combining seemingly haphazardly in both domains: he describes the mixed type of diction, for instance, as ‘compounded’ (σύνθετος ) but also as ‘well-blended’ ( εὔκρατος ). 126 Still, he clearly prefers the former interpretation, for when he turns to specific examples of mixed diction, he builds exclusively on the concept of juxtaposition: he sharply demarcates simple passages from elevated ones, separating them through radical breaks. 127 Dionysius regards simple and elevated diction predominantly as units that are not used simultaneously, but in alternation. In word arrangement, by contrast, Dionysius prefers the concept of blending. He does not describe the extreme harmonies as blocks that can only be used in their entirety, but he assumes that one can be diluted with the other: he repeatedly asserts that mixed composition is formed ‘through the relaxation and the intensification of the extremes’ ( κατὰ τὴν ἄνεσίν τε καὶ ἐπίτασιν τῶν ἄκρων ). 128 Thus, the mixed harmony can be fine-tuned to become either we will see in the remainder of the present section, this interpretation works well with respect to the mixed type of diction, but not with respect to the mixed type of arrangement. Bonner (1938) 261–263, conversely, thinks that the middle style ‘is formed not by combination but by a process of selection of the best points in the two extremes and avoidance of the excesses’. Yet, this interpretation is more apposite to Dionysius’ theory of diction than to his theory of arrangement. We must conclude, therefore, that Dionysius uses both concepts of mixture in his understanding of stylistic mixtures: he is less concerned with the consistency than with the expediency of his theoretical concepts. Cf. also Martinho (2010) for a discussion of Dionysius’ conceptions of mixing. 126 Dion. Hal. Dem. 3.1, 3.5. 127 Cf. Pohl (1968) 31–32. The analysis of passages from Pl. Phdr. in Dem. 7 exemplifies this approach. Dionysius states that the language in the first part of the dialogue (227a–236e) excels in the essential virtues of style: the text ‘has much grace and is full of charm’ ( πολλὴν ὥραν ἔχει καὶ χαρίτων ἐστὶ μεστά ). Dionysius argues that Plato turns to the additional virtues, as soon as the reading of Lysias’ speech begins (237a): ‘Like a violent wind bursting out of the calm still air, he shatters the purity of expressions by resorting to tasteless artificiality’ ( ὥσπερ ἐξ ἀέρος εὐδίου καὶ σταθεροῦ πολὺς ἄνεμος καταρραγείς , ταράττει τὸ καθαρὸν τῆς φράσεως ἐς ποιητικὴν ἐκφέρων ἀπειροκαλίαν ). Other such analyses of individual passages can be found at Dem. 12–14 (on Dem. Con. 3–9), ibid. 17–20 (on Isoc. Pac. 41–50), ibid. 21–22 (on Dem. Olynth. III 23–32), ibid. 24–30 (on various passages from Pl. Menex. 236a–248c), and ibid. 31–32 (on Dem. De cor. 199–208). 128 The image of ‘relaxation and intensification’, which refers to the strings of the lyre (cf. Pl. Resp. 349e), is one of Dionysius’ favorite images to describe the mixed style of word arrangement: see Comp. 21.5, 24.3; Dem. 37.1, 44.1 and 46.2. He also applies it to Demosthenes’ ‘intensity’ ( τόνος ) in Dem. 13.10, and to Isocrates’ lack thereof in Isoc. 13.4; cf. Philod. Rhet. 4 col. 17.1–9 p. 198 Sudhaus. Dionysius does not, however, apply the image to the mixed type of word selection. At Comp. 21.5, he warns us not to take the comparison with music too seriously: ‘It is not the same as in music, where the middle note is equidistant from the lowest and the

136

THE PLAIN , THE GRAND AND THE IN-BETWEEN smoother or rougher according to the relative proportion of smooth and rough elements in the mix: the arrangement becomes rougher, for instance, through the inclusion of hiatus and slow rhythms, and it becomes smoother with the addition of fluid word combinations and quicker rhythms (section 4.4). In any case, Dionysius describes the resulting style as the product of a process of ‘selection’ ( ἐκλογή ) of those elements that are ‘most useful’ ( χρησιμώτατα ) and ‘most powerful’ ( κράτιστα ) for the author’s purposes. 129 Thus, mixed styles come in many varieties: those authors who have used the mixed style ‘have not all studied the same aspects of it or treated them in the same way, but some have studied one set more, others another. Moreover, when authors use the same aspects, some have intensified ( ἐπέτειναν ) them, some have relaxed ( ἀνῆκαν ) them in different ways.’ 130 Regardless of the precise interpretation of Dionysius’ concept of mixing, the fact that the Greek teacher interprets the middle style as a mixture allows him to focus on two stylistic qualities that he considers paramount features of literary excellence: we already encountered these features in Dionysius’ assessment of the genius of Demosthenes (section 2.3.3) and of the inadequacy of Hegesias of Magnesia (section 2.4.3). First of all, the idea of a mixture highlights the importance of stylistic versatility. Dionysius measures the success of the mixed style in terms of ‘diversity’ ( ποικιλία ), ‘variation’ ( μεταβολή ), ‘timing’ ( εὐκαιρία ), ‘balance’ (συμμετρία ) and ‘appropriateness’ ( πρέπον ). 131 He praises Homer’s composition, because he is the poet ‘with the most voices’ ( πολυφωνότατος ) and because ‘he has pushed his varied mixture to the limit’ ( εἰς ἄκρον διαπεποίκιλται ). 132 Dionysius devotes special attention to the author’s ability to seize ‘the right moment’ ( καιρός ): Plato and Isocrates fall short because

highest ; the middle style in literature does not in the same way stand at an equal distance from the two extremes’ (οὐ γὰρ ὥσπερ ἐν μουσικῇ τὸ ἴσον ἀπέχει τῆς νήτης καὶ τῆς ὑπάτης ἡ μέση , τὸν αὐτὸν τρόπον καὶ ἐν λόγοις ὁ μέσος χαρακτὴρ ἑκατέρου τῶν ἄκρων ἴσον ἀφέστηκεν ). Mixed harmony can indeed at times be closer to rough harmony, and at other times border on the smooth type: see also Aujac and Lebel (1981) 218. 129 Dion. Hal. Dem. 8.2 (on word selection), and ibid. 41.1, Comp. 24.1 (on word arrangement). 130 Dion. Hal. Comp. 24.3: Οἵ τε χρησάμενοι αὐτῇ οὐ τὰ αὐτὰ πάντες οὐδ’ ὁμοίως ἐπετήδευσαν , ἀλλ ’ οἳ μὲν ταῦτα μᾶλλον , οἳ δ’ ἐκεῖνα , ἐπέτεινάν τε καὶ ἀνῆκαν ἄλλως ἄλλοι τὰ αὐτά . The pronoun αὐτῇ refers to the mixed type of arrangement ( μικτὴ ἁρμονία ). 131 Dion. Hal. Dem. 34.5 lists these items, with the exception of variation ( μεταβολή ), as the virtues that Demosthenes exhibits more than any other exponent of the mixed type of diction. Dion. Hal. Comp. 11.1 highlights the importance of ‘variation’ ( μεταβολή ) and ‘appropriateness’ ( πρέπον ) in word arrangement: they are two of the four means to attain beauty and pleasure, in addition to ‘rhythm’ ( ῥυθμός ) and ‘tone’ ( μέλος ). Cf. section 4.4 n. 66. 132 Dion. Hal. Comp. 16.8 and 24.4. The translation is mine.

137

CHAPTER THREE their words are sometimes ‘ill-timed’ ( ἄκαιρος ), but Demosthenes always ‘aims for what is needed and makes his style conform to the occasion’ ( τοῦ ἀρκοῦντος στοχάζεται καὶ τοὺς καιροὺς συμμετρεῖται ). 133 Like Demosthenes, the consummate author of the mixed style is able to adapt his discourse to any given situation. Secondly, Dionysius’ notion of the middle style as a mixture of two extremes permits him to advertise a style that is perfectly suited for practical oratory in the public assemblies and the law-courts: as we already saw, Dionysius celebrates the practical force of Demosthenes’ speeches as one of the orator’s prime virtues (section 2.3.3). The critic reasons that the people who attend judicial and political meetings ‘are neither all outstanding intellectual geniuses like Thucydides, nor all laymen with no experience of how a good speech is composed’.134 Dionysius’ ideal orator consequently adapts his choice of words to both groups: the connoisseurs and savants should be addressed in the striking, elevated diction of a Thucydides, whereas the simple words of a Lysias appeal to the uneducated majority. In a word, a mixed audience requires a mixed style: ‘The speech which aims to persuade both these extreme classes of audience is less likely to fail in its objective.’ 135 According to Dionysius, Demosthenes has understood this principle perfectly: his style is ‘most perfectly adapted to all aspects of human nature’ ( πρὸς ἅπασαν ἀνθρώπου φύσιν ἡρμοσμένη ) and it has ‘universal appeal’ ( κοινή τε καὶ φιλάνθρωπος ). 136 In the next section, I will further explore the consequences of Dionysius’ ideas about mixed audiences, and compare his views to the ones that Cicero articulates. Finally, the Greek critic conjures up the world of Homer’s epics to convey the overwhelming power that the mixed style can wield over its readers and listeners. Much like the shapeshifting gods and heroes from mythology, the master of the mixed style is able to

133 The term καιρός is notoriously difficult to grasp or translate: as a rhetorical-technical term, Trédé (1992) 247– 253 defines it as ‘le principe qui gouverne le choix d’une argumentation, les moyens utilisés pour prouver et, plus particulièrement, le style adopté’. Kennedy (1963) 66 shows the close relationship between the term καιρός and other rhetorical concepts, such as ‘appropriateness’ ( πρέπον ), ‘propriety’ ( οἰκειότης ) and ‘due measure’ (μέτρον ). On failures to observe καιρός , see e.g. Dion. Hal. Dem. 4.4 (Isocrates) and ibid. 5.5 (Plato); on successes, see e.g. ibid. 10.3 (Demosthenes). 134 Dion. Hal. Dem. 15.2: Οὔτε δεινοὶ καὶ περιττοὶ πάντες εἰσὶ καὶ τὸν Θουκυδίδου νοῦν ἔχοντες οὔθ’ ἅπαντες ἰδιῶται καὶ κατασκευῆς λόγων γενναίων ἄπειροι . For Dionysius’ views on the audiences of oratory in the classical Greek past and in the Roman present, see section 3.6 below. 135 Dion. Hal. Dem. 15.6: Ὁ δὲ ἀμφότερα τἀκροατήρια πείθειν ζητῶν ἧττον ἀποτεύξεται τοῦ τέλους . The word λόγος should be supplied as the subject of the verb ἀποτεύξεται . 136 Dion. Hal. Dem. 33.1, 33.3.

138

THE PLAIN , THE GRAND AND THE IN-BETWEEN present himself to his audience in whichever way he desires. Dionysius compares Demosthenes’ mixed type of diction to the ever-changing sea-god Proteus: the orator ‘has created a single style out of many (...) which has a character not at all unlike that of Proteus as portrayed by the mythological poets, who effortlessly assumed every kind of shape, being either a god or superhuman, with the power to deceive human eyes’.137 Similarly, we already saw that Dionysius adduces the image of the goddess Athena to describe the awesome power of word arrangement (section 2.4.3): ‘It seems to me that one would not be wrong to compare word arrangement to Athena in Homer, for she used to make the same Odysseus appear in different forms at different times.’138 Τhese comparisons show that the author who successfully applies the mixed style puts a spell on his audience: like Proteus and Odysseus he can change into many forms, thus dictating precisely how he is perceived by the audience. In conclusion, Dionysius’ interpretation of the stylistic middle makes a rather ambiguous impression at first sight: he not only builds his views on the Peripatetic theory of the ‘mean’ ( μεσότης ), but he also presents his middle style as a mixture, or indeed a compound, of opposite styles. All in all, he does not seem to be overly worried about the apparent inconsistence of these ideas, as long they contribute to his mission of defining the perfect style. Dionysius exploits the concepts of mean, middle and mixture to serve his argument in the best way possible: in the end, he is able to define his preferred style as a versatile style with universal appeal, which can be applied to all circumstances and which caters for every audience.

3.6 Greek and Roman Audiences: the Three Styles in Action In this last section of this chapter, I will take a closer look at the audiences that Cicero and Dionysius conjure up in their discussions of oratorical style. While the two authors interpret the three-style formula, as we have seen, in divergent ways, they share a general idea about

137 Dion. Hal. Dem. 8.2–3: Μίαν ἐκ πολλῶν διάλεκτον ἀπετέλει (...) οὐδὲν διαλλάττουσαν τοῦ μεμυθευμένου παρὰ τοῖς ἀρχαίοις ποιηταῖς Πρωτέως , ὃς ἅπασαν ἰδέαν μορφῆς ἀμογητὶ μετελάμβανεν , εἴτε θεὸς ἢ δαίμων τις ἐκεῖνος ἄρα ἦν παρακρουόμενος ὄψεις τὰς ἀνθρωπίνας . Proteus’ ability to change his appearance features in Hom. Od. 4.455–458, Verg. G. 4.405–414, 440–442 and, presumably, in Aeschylus’ lost satyr play Proteus. Dionysius adopts an allegorical interpretation of the figure, not considering him a real god or hero, but rather a ‘creature that stands for the varied language of a clever man, beguiling every ear’ ( διαλέκτου ποικίλον τι χρῆμα ἐν ἀνδρὶ σοφῷ, πάσης ἀπατηλὸν ἀκοῆς). I do not agree with Costil (1949) 473 and Aujac and Lebel (1981) 165, who think that Dionysius must be drawing on Stoic ideas, because he uses an allegorical interpretation of Homer. 138 Dion. Hal. Comp. 4.12: Καὶ μοι δοκεῖ τις οὐκ ἂν ἁμαρτεῖν εἰκάσας αὐτὴν τῇ Ὁμηρικῇ Ἀθηνᾷ· ἐκείνη τε γὰρ τὸν Ὀδυσσέα τὸν αὐτὸν ὄντα ἄλλοτε ἀλλοῖον ἐποίει φαίνεσθαι . Cf. section 2.4.3 n. 145 above.

139

CHAPTER THREE the practical application of the theory: both men associate their ideal orator with real-life oratory rather than with epideictic showpieces. In their view, the versatile repertoire of the three styles fulfills its true potential in the lawsuits of the courts and in the deliberations of the political assemblies, where the stakes are really high.139 Yet, the settings that Cicero refers to are not quite the same as those that Dionysius has in mind: the Roman deals with orations that are to be performed in front of the Roman crowds of the Forum, whereas the Greek seems to evoke speeches that are to be delivered in front of the Greek masses of the agora. This disparity underlies a fundamental difference in their appreciation of two out of the three styles: they disagree considerably about the proper application of plain and grand oratory. These considerations raise a fundamental question: does a Roman audience ask for a different approach than a congregation of Greek listeners? That Cicero designs his theory of three styles for use on the Roman Forum is obvious from his frequent references to the religious, administrative, legal and social center of the ancient city. He describes his ideal orator as ‘the man who is able to speak on the Forum and in public trials ( in foro causisque civilibus ) so as to prove, to please and to sway’. He attributes the greatest power to the grand style, because ‘it is the one thing of all that avails most in winning verdicts ( ad obtinendas causas )’. Throughout his Orator , Cicero is only interested in ‘real cases’ ( verae causae ) and in the ‘struggles of the Forum’ ( forenses contentiones ). 140 Hence, he virtually glosses over the intermediate style, as we have seen, because it was not indigenous to the Forum, springing from the ‘sources of the sophists’ (sophistarum fontes ).141 Likewise, he advises his orator against imitating the styles of philosophers, historians and epideictic speakers, as their styles are not readily usable ‘for public use on the Forum’ ( ad forensem usum et publicum ) or ‘in the reality of public cases’ ( in veritate causarum ). 142 For Cicero, then, the Roman Forum stands for the gritty, cutthroat reality, in which true eloquence can shine.

139 The opposition between epideictic oratory on the one hand and judicial and deliberative oratory on the other hand is already present in the distinction between the ‘written style’ ( λέξις γραφική ) and the ‘performative style’ (λέξις ἀγωνιστική ) in Arist. Rh. 3.12: see esp. Innes (2007) 151–156 and Ooms and De Jonge (2013) 100–101. 140 Phrases such as the ones quoted above permeate Cic. Orat. For a selection of passages, see ibid. 12, 30, 37, 38, 63, 69, 120, 208, 209, 221, 225. 141 Cic. Orat. 96. Cf. Brut. 37, which discusses the oratory of Demetrius of Phalerum, Cicero’s prime exponent of middle-style oratory, in a similar vein: he is said to have emerged into the ‘heat and dust of the action’ ( in solem et pulverem ) ‘from the comfortable shadows’ ( ex umbraculis ) of Theophrastus’ school, not ‘from a soldier’s tent’ ( e militari tabernaculo ). Cf. section 3.2.2 above, esp. n. 45–46. 142 See Cic. Orat. 37–38, 62–64 (on epideictic oratory), 39–42, 66 (on historiography), and 65 (on philosophy).

140

THE PLAIN , THE GRAND AND THE IN-BETWEEN

Like his Roman predecessor, Dionysius prefers high-stakes action over detached leisure: according to the Greek critic, ‘contests’ ( ἀγῶνες ) and the concomitant ‘contest- speeches’ ( ἐναγώνιοι λόγοι ) constitute the ideal breeding ground of oratorical excellence. Specifically, Dionysius refers to speeches delivered before ‘the public assemblies (ἐκκλησίαι ), the courts ( δικαστήρια ) and other meetings where there is need of civic speeches (πολιτικοὶ λόγοι )’. 143 Unlike Cicero, however, Dionysius does not seem to have the Roman Forum in mind, when he refers to these institutions. Rather, he appeals particularly to Greek democracy and patriotism: he is still overwhelmed, for instance, when he picks up the rant against king Philip in Demosthenes’ Third Olynthiac , and he becomes a ‘patriotic supporter of democracy’ ( φιλόπολίς τε καὶ φιλόδημος ), when he reads the antibarbarian rhetoric of Isocrates’ Panegyricus . Apparently, Dionysius especially admires those speeches that, centuries after their initial delivery, still incite philhellenic sentiments in their audience. 144 This does not, of course, preclude the possibility that Dionysius addresses the orators of the Forum: after all, the Roman aristocrats Metilius Rufus and Q. Aelius Tubero are among the addressees of his critical essays. Still, it is incontestable that the critic only discusses Greek literary works in their Greek context.145 Incidentally, Metilius Rufus had ample opportunity to put Dionysius’ teachings about Greek rhetoric into practice: in his political career, he functioned as proconsul of Achaea and, possibly, as legate of Galatia. 146

143 Dion. Hal. Dem. 15.2: Αἱ ἐκκλησίαι καὶ τὰ διακαστήρια καὶ οἱ ἄλλοι σύλλογοι , ἔνθα πολιτικῶν δεῖ λόγων . For the use of ἀγών and cognates to refer to speeches in the context of debate, that is, to judicial and deliberative oratory, see Ooms and De Jonge (2013) 100–102. Dion. Hal. Dem. 18.5 refers to the speakers in such situations as ‘contestants’ ( ἀθληταί ), who need a ‘firm grip’ ( ἰσχυραὶ ἁφαί ) and an ‘ineluctable hold’ ( ἄφυκτοι λαβαί ). On Dionysius’ conception of ‘civic discourse’ (πολιτικὸς λόγος ), see n. 144 below. 144 See Dion. Hal. Dem. 22.4–7 (cf. section 2.2 n. 21 above) and Isoc. 5.1–4. Wiater (2011) adduces these passages, when he claims that Dionysius’ classicism is a ‘model for Greek cultural identity’, which aims to separate Greek and Roman audiences: this dissertation, esp. section 1.2 above, objects to this view on account of the numerous connections between the views of Dionysius and Cicero: cf. section 5.5.2 below on the praise of Rome in Ant. orat. 3.1–3. We should note that Isoc. Paneg. is not a ‘contest-speech’ ( λόγος ἐναγώνιος ) but an epideictic speech: cf. n. 139 and 143 above. Yet, as Wiater (2011) 65–77 points out, the oration counts as a ‘civic discourse’ ( πολιτικὸς λόγος ), which Dion. Hal. Dem. 15.2 places in the same category as judicial and deliberative speeches: according to Dionysius, πολιτικοὶ λόγοι are all discourses that embody the civic virtues of Classical Greece. 145 On the Greek and/or Roman context of Dionysius’ critical treatises, see section 1.2 above. On his Greek and Roman addressees, see section 1.5 above. 146 See Bowersock (1965) 132 n. 2, cf. section 1.5 n. 91 above. Goudriaan (1989) 29–37 shows that the old institutions of Classical Greece, viz., the people’s assembly ( ἐκκλησία ), the council ( βούλη ) and the lawcourts

141

CHAPTER THREE

What do our sources have to say about the make-up of the crowds that flock together into the Roman Forum and the Greek agoras respectively in order to participate in the various public gatherings? Both Cicero and Dionysius touch on this topic in their critical works: interestingly, they divide the listeners, Roman or Greek, roughly into the same two groups. Cicero, for one, notes that the orator has to deal with the ‘judgment of experts’ ( iudicium intellegentium , iudicium sapientium ) as well as with the ‘judgment of the crowd’ ( iudicium vulgi ) which consists largely of ‘untrained people’ ( imperiti ). 147 Likewise, Dionysius distinguishes between a minority of ‘outstanding intellectual geniuses’ ( δεινοὶ καὶ περιττοί ), who are ‘well-versed in public life, having gone through a broad education’ ( οἱ πολιτικοί τε καὶ ἀπ’ ἀγορᾶς καὶ διὰ τῆς ἐγκυκλίου παιδείας ἐληλυθότες ) and a majority of ‘laymen who have no experience of how a good speech is composed’ ( ἰδιῶται καὶ κατασκευῆς λόγων γενναίων ἄπειροι ). 148 Cicero and Dionysius often refer to this distinction between laymen and experts: as we will see in the chapter about word arrangement, they both argue that the ignorant masses rely on the ‘irrational criterion of perception’ ( ἄλογος αἴσθησις , tacitus sensus ), while the experts can fall back on a theoretically underpinned rational judgment, when they evaluate the quality of artistic prose (sections 4.1 and 4.3). Thus, when an orator delivers a public speech before a large crowd of people—be they Romans or Greeks—he has to take into account both the uneducated masses and the cultural elite. According to Cicero and Dionysius, the orator can tackle this problem by drawing rhetorical devices from various stylistic registers, using the convenient repertoire of the three styles. Dionysius addresses this topic, as we have seen, in his discussion of the three types of

(δικαστήρια ), were still active during the Hellenistic and Roman eras, be it that they came to function as obsolete symbols of a free democracy that no longer existed: cf. also Heldmann (1982) 98–122, who adopts a pessimistic view about the state of eloquence in Dionysius’ day. Yet, on the basis of Acts 19:23–40, Goudriaan shows that rhetoric was still very much alive in the cities of the Greek East: thus, Dionysius’ focus on judicial and deliberative oratory was by no means purely nostalgic, but it could serve a practical purpose in his own days. 147 Cic. Brut. 183–200 argues that the judgment of the ignorant crowd always coincides with the judgment of the literary experts: in his view, both groups can instinctively distinguish between good and bad oratory, although only the latter group is capable of substantiating its opinions. The whole discussion prepares for Cicero’s attack against the self-styled Atticists in Brut. 283–291: according to the rhetorician, Calvus and his followers are content with pleasing a handful of experts, neglecting the judgment of the masses. Cf. De or. 3.195–199. Schenkeveld (1988) conveniently summarizes Cicero’s discussions of the ‘judgment of the crowd’ ( iudicium vulgi ) in De or. and Brut. , showing that the Roman critic draws on the Greek concept of ‘irrational perception’ (ἄλογος αἴσθησις ): see esp. section 4.1 n. 5 and section 4.4 n. 56 and 62 below. 148 Dion. Hal. Dem. 15.2–4. Cf. section 3.5 above for a discussion of Dionysius’ insistence on a universal style.

142

THE PLAIN , THE GRAND AND THE IN-BETWEEN diction, advocating a judicious mixture of elevated and plain vocabulary: he explains that simple, everyday words appeal to the common folk, whereas extravagant, striking vocabulary pleases people of refinement (section 3.5). Cicero, however, tells a rather different story: according to him, the plain style appeals to the elite, not to the masses. The Roman author reports that the self-styled Atticists, fervent champions of simple, unadorned oratory, receive ‘critical acclaim’ ( approbationes ) for their stylistic frugality and that they enjoy a reputation of wisdom ( prudentia ) on this account.149 He associates the grand style, conversely, with the unschooled crowd: Cicero argues that the Atticists might be able to earn the ‘assent of the few’ ( paucorum approbatio ), but that they will never acquire the ‘admiration of the masses’ (assensus vulgi ). In order to win over the crowd, a richer, more ornate style is required, or, in Cicero’s words: ‘The benches call for a louder and fuller voice.’ 150 Concerning their appreciation of simple and elevated oratory, Cicero and Dionysius seem to be diametrically opposed to each other. Their disagreement can be explained on two accounts. To start with, it is the inevitable outcome of their divergent interpretations of the extreme types of diction (section 3.2). As we have seen, Dionysius explains the contrast between the simple and the elevated style as one between ‘familiar’ ( συνήθης ) and ‘natural’ (φυσικός ) language on the one hand, and ‘extraordinary’ ( ἐξηλλαγμένος ) and ‘strange’ (ξενός ) language on the other hand (tables 2 and 5): laymen, whose literary expertise is limited, naturally feel more comfortable with the former, whereas experts, who like to be challenged intellectually, are more pleased with the latter. Cicero, conversely, describes the

149 Cic. Orat. 236: ‘To express my opinion briefly, the fact of the matter is that to speak with well-knit rhythm without ideas is folly, to present ideas without order and rhythm in the language is to be speechless; but such a kind of speechlessness that those who use it could be considered, not stupid, but on the whole wise ( res se autem sic habet, ut brevissime dicam quod sentio: composite et apte sine sententiis dicere insania est, sententiose autem sine verborum et ordine et modo infantia, ut ea qui utantur non stulti homines haberi possint, etiam plerumque prudentes ). In the same passage, Cicero adds that the perfect orator should not merely aim for ‘approval’ ( approbationes ) but also for ‘admiration, cheers and applause’ (admirationes, clamores, plausus ): the latter can, in his view, only be incited by an orator who is prepared to use every stylistic trick in the book: cf. section 2.3.2 above on Cicero’s discussion of Demosthenes’ success in his use of the grand style. 150 For the appreciation of the Atticists’ rhetoric by experts and laymen, see Cic. Brut. 191. See also ibid. 289: Subsellia grandiorem et pleniorem vocem desiderant . Hendrickson in Hendrickson and Hubbell (1962) 252 notes that the presence of ‘benches’ ( subsellia ) indicates a large trial; in private cases, the parties would remain standing. Cicero ridicules the Atticists’ failures to please large crowds: ‘When these Atticists of ours speak, they are deserted not only by the curious crowd, which is humiliating enough, but even by the friends and supporters of their clients’ ( at cum isti Attici dicunt, non modo a corona, quod est ipsum miserabile, sed etiam ab advocatis relinquuntur ).

143

CHAPTER THREE plain style as ‘astute’ ( acutus ), designed to ‘prove’ ( probare ), whereas he characterizes the grand style as ‘passionate’ ( vehemens ), aiming to ‘sway’ (flectere ) the audience (tables 1 and 3): the connoisseurs who build their judgment on rational analysis feel at home in the former style, but the masses, who rely on their instincts, are sensitive to impassioned orations. Needless to say, Cicero’s insistence that the grand style is the most powerful in winning big cases in front of large crowds, corroborates the defense of his own opulent style against the sustained objections of his Atticist opponents. In addition, the disagreement between Cicero and Dionysius reveals a difference between Greek and Roman audiences in the first century BC: each group seems to react rather differently to classicizing prose, which draws on the works of literary masters who had been dead for centuries. It is mind-boggling to realize that Cicero, Dionysius and their contemporaries advocated the imitation of authors who were at least as old to them as William Shakespeare is to modern speakers of English: would the participants in a creative writing course today be advised to imitate the Bard’s timeworn stylistic intricacies? For Greek audiences, this issue was not as pressing as for their Roman coevals: even after so many years, the straightforward style of Lysias was still more accessible than the convoluted sentences of Thucydides. Therefore, Dionysius can commend the imitation of the former in addressing the hoi polloi, while he urges those who aim to appeal to literary specialists to study the latter. It is not surprising that Dionysius—as well as, no doubt, his fellow Greek scholars—considered Lysianic language to be suitable ‘for practical use’ ( εἰς χρῆσιν ), while they regarded Thucydidean prose as particularly apt ‘for permanent literary value’ ( εἰς ἀνάθημα καὶ κτῆμα ).151 For Cicero and the native speakers of Latin, however, the age-old works of Lysias, Thucydides and the other exponents of Classical Greek literature might have been equally hard to understand. Hence, the Roman Atticists’ predilection for Lysias is not born from a desire to appeal to large crowds of unlettered people: rather, we will see that they chose Lysias as their preferred model, because they saw in his prose the embodiment of Attic purity, masculinity and moderation (section 5.6). Cicero, on the other hand, warns his readers that such monuments for Classical Attic orators might please the literature buffs of Rome, but that these works are ultimately unsuccessful on the Forum. 152 He tells a cautionary tale about the

151 Dion. Hal. Dem. 10.3. The passage mockingly alludes to Thuc. 1.22.4: ‘My work has been composed, not as a prize-essay to be heard for the moment, but as a possession for all time’ ( κτῆμά τε ἐς αἰεὶ μᾶλλον ἢ ἀγώνισμα ἐς τὸ παραχρῆμα ἀκούειν ξύγκειται ). 152 For the failure of the Atticists to please large crowds, see n. 150 above.

144

THE PLAIN , THE GRAND AND THE IN-BETWEEN poet Antimachus of Colophon, who, in the midst of reciting an obscure poem, noticed that all his listeners had left except for Plato, whereupon he remarked: ‘I shall go on reading just the same; for me, Plato alone is as good as a hundred thousand.’ 153 In Cicero’s view, his Athenophile opponents alienate their audience by committing themselves with relentless zeal to the imitation of an antique literary model. 154 Whereas imitations of Lysias came across as plain and straightforward to Greeks, they could be condemned as elitist—or, alternatively, praised as sophisticated—by Romans. A Roman audience, in conclusion, indeed calls for a different approach than a Greek audience: the imitation of the Greek models for the plain and grand stylistic registers had different effects on the crowds of the Roman Forum than on the masses of the Greek agoras. Apparently, the sober style of a Lysias was music for the millions among the Greeks, whereas it catered for a select group of intellectuals among the Romans. An opulent style, conversely, could be regarded as pretentious in the Greek courts and assemblies, and as crowd-pleasing in the meetings of Roman institutions.

3.7 Conclusion This chapter has explored the notion, articulated by several Greek and Roman authors, that there exist three types of style, or stylistic registers. We have seen that this idea is integral to the stylistic discourse of Late-Republican and Augustan Rome: indeed, such threefold divisions are not attested before the first century BC, and later sources seem to retain them solely on the authority of Cicero and Varro. Attempts by modern classicists to trace the origins and early development of the three-style formula have been unconvincing, simply because there did not exist a uniform doctrine of three styles. Rather, the extant articulations of the triple scheme show that the three styles provide a flexible framework that could be interpreted in various ways according to the specific purposes, preferences and programs of its users. Hence, we have seen that Cicero, Dionysius, Varro and the author of Rhetorica ad

153 Cic. Brut. 191: Legam nihilo minus; Plato enim mihi unus instar est centum milium. Antimachus probably flourished during the Peloponnesian War, he authored an epic poem Thebais , in at least five volumes, and an elegiac poem Lyde , and he was famous for his predilection for glosses: see Matthews (1996) 64–76. The anecdote about Plato’s approval and the rejection of the public is also related by Plut. Lys. 18. 154 Cf. section 4.5 below for Cicero’s view on the development of oratorical style: according to him, the orators of his day had a formidable arsenal of stylistic devices at their disposal: speakers who chose not to avail themselves of it, ran the risk of alienating their audience. For the religious zeal of Calvus and his fellow- Atticists, see section 5.6.1 below.

145

CHAPTER THREE

Herennium use a diverse palette of nouns and adjectives to define each of their three styles. Moreover, this chapter has shown that Cicero and Dionysius each agilely connect their stylistic triad to several of their other rhetorical doctrines, and that they put it to use to make sense of the overwhelming literary legacy of Classical Greece. In the hands of our Greek and Roman authors, the theory of three styles is a versatile tool with an enormous potential. More than anything else, the authors use the malleable three-style categorization as a generous source for arguments in their rhetorical and critical treatises. For Cicero, the triple scheme provides him with several expedient arguments in his polemic with the self-styled Attic orators who rose to prominence in the late 50s BC; his discussion focuses specifically on the simple and grand styles. Dionysius, conversely, concentrates on the concept of the intermediate style, which he sometimes presents as a sort of Aristotelian mean, sometimes as a mixture of the two extreme styles: we have seen that the Greek critic is not worried about the consistency of his doctrines, as long as they serve his purpose of defining the perfect, all- round and universal type of style. Lastly, the stylistic discussions of Cicero and Dionysius show that Greek audiences respond differently to the grandeur and simplicity of classicizing prose than their Roman counterparts: what is mainstream among Greeks, is considered niche by Romans, and vice versa. The next chapter will build on the foregoing discussion of the three styles: I will focus on the topic of word arrangement, which Dionysius, as we have seen, divides in a rough, a smooth and a mixed type. Specifically, I will focus on the first of these categories: notwithstanding the standard view that classical literature is full of agreeable charm, I will be looking for its jarring harshness.

146

Chapter 4 OFFENDING THE EARS : GREEK AND LATIN VIEWS ON ROUGH WORD ARRANGEMENT

4.1 Introduction Like the doctrine of three styles, the arrangement of words (in Greek σύνθεσις τῶν ὀνομάτων , in Latin compositio verborum ) is a hot topic in the stylistic discourse of Late-Republican and Augustan Rome. There are roughly two ancient approaches to word arrangement, or ‘composition’, as the topic is usually referred to in modern scholarship. 1 First, there is the logical-syntactical perspective, which evaluates word order on the basis of grammatical rules, e.g. ‘nouns should precede verbs’ or ‘verbs should precede adverbs’. 2 In his monograph On the Arrangement of Words (περὶ συνθέσεως ὀνομάτων ), Dionysius of Halicarnassus explores and ultimately rejects this approach, because ‘at times the arrangement became charming and beautiful by these and similar principles, but at other times not by these, but by the opposite sort.’ 3 Instead, Dionysius adopts an alternative perspective that is better suited to evaluate

1 The modern usage of the term ‘composition’, however, has a much wider application: cf. ‘the action or act of producing a creative work such as a poem or piece of music’ ( OED s.v. 2.1). The term can also give the false impression that it is related to the arrangement of subject matter, which can be avoided by using ‘literary composition’, ‘stylistic composition’ or ‘verbal composition’. I recognize that the designation ‘word arrangement’ is also somewhat imperfect (i.e., it does not account for the attention that our sources pay to arranging letters, syllables, clauses and sentences), but it aptly brings out the meaning of the ancient terminology, which consists of the nouns σύνθεσις /compositio (a process of ‘putting together’ or ‘arranging’) and ὀνόματα /verba (‘words’). Cf. Dion. Hal. Comp. 2.1, who defines σύνθεσις τῶν ὀνομάτων as ‘a certain process of arranging the parts of speech’ ( ποιά τις θέσις παρ ’ ἄλληλα τῶν τοῦ λόγου μορίων ). 2 Dion. Hal. Comp. 5 discusses these and six similar syntactical rules: the critic presents the logical-syntactical approach as a ‘natural starting point’ ( φυσικὴ ἀφορμή ) for his investigation of word arrangement. De Jonge (2008) 273–315 shows that Dionysius’ discussion of natural word order is largely inspired by Stoic works on language, e.g., Chrysippus’ On the Order of the Parts of Speech (Περὶ συντάξεως τῶν τοῦ λόγου μερῶν), quoted by Dion. Hal. Comp. 4.17, 4.20. Other texts which apply a logical-syntactical approach to word order are Demetr. Eloc. 199–201, Long. Subl. 22.1 and Quint. Inst. orat. 9.4.23–27: see De Jonge (2001) and (2008) 315– 327. Cf. n. 4 below on Horace’s ‘clever combination’ ( callida iunctura ). 3 Dion. Hal. Comp. 5.10: Τοτὲ μὲν γὰρ ἐκ τούτων ἐγίνετο καὶ τῶν ὁμοίων αὐτοῖς ἡδεῖα ἡ σύνθεσις καὶ καλή , τοτὲ δ’ ἐκ τῶν μὴ τοιούτων ἀλλ ’ ἐναντίων . Dionysius’ rejection of the logical-syntactical approach to word arrangement in Comp. does not preclude his use of grammatical terminology throughout the work: cf. his references to the theory of the parts of speech at Comp. 2.1, 6 and 22–24, which are discussed by De Jonge (2008) 183–213. I agree with De Jonge (2008) 47 that Dionysius’ use of grammatical teachings does not make

147

CHAPTER FOUR charm and beauty in word arrangement: this second approach, which can be described as musical-aesthetic, focuses on the sound that supervenes on the word order, or, more accurately, on the audience’s aural perception ( αἴσθησις , sensus ) of tone ( μέλος , sonus ) and rhythm ( ῥυθμός , numerus ). The latter approach is dominant in the stylistic discourse of Late- Republican and Augustan Rome. The aural evaluation of word arrangement features prominently in most extant stylistic discussions of prose and poetry from the second century BC onwards: it is not only attested in Greek sources (Demetrius, Philodemus, Dionysius, Longinus), but in Latin texts as well (Lucilius, Rhetorica ad Herennium , Cicero, Quintilian). 4 The shared Greek and Latin discourse on word arrangement, which emerges from these texts, is built on the assumption that the judgment of word order ultimately resides not in the mind, but in the ear (ἀκοή , auris ), which bases its assessment not on logical principles, but on the ‘irrational criterion of perception’ ( ἄλογος αἴσθησις , tacitus sensus ). 5 According to the authors, the acoustic effects of word order can be divided into two categories: some collocations produce sounds that ‘please the ear’ ( γλυκαίνειν τὴν ἀκοήν , permulcere auris ), while others produce sounds that ‘grate the ear’ ( πικραίνειν τὴν ἀκοήν , offendere auris ). 6 Dionysius’ threefold system of

his work a grammatical treatise: ‘Dionysius combines a wide knowledge of many different disciplines on the one hand with a focus on the practical purposes of his own work on the other hand.’ 4 Demetr. Eloc. esp. 38–74, 179–184, 204–208, 241–271; Philod. Poem. passim; Dion. Hal. Dem. 36–49, Comp. passim; Long. Subl. 39–42; Lucil. fr. 74–76, 367–370, 1188 Krenkel (= fr. 84–86, 389–392, 419 Warmington); Rhet. Her. 4.18; Cic. De or. 3.171–199, Orat. 149–236; Quint. Inst. orat. 9.4. This chapter will not pay much attention to the notion of ‘clever combination’ ( callida iunctura ) in Hor. Ars. P. 46–48 and 240–243, for the poet does not refer to tone or rhythm, but rather to syntax and semantics: in his view, a poet can add distinction to his style, ‘if a clever combination makes a familiar word new’ ( notum si callida verbum reddiderit iunctura novum ). Cf. Brink (1971) 139. Yet, De Jonge (2019a) points out four striking parallels between the views of Dionysius and Horace about word arrangement: both men stress the importance of commonplace words, clever skill, metamorphosis and distinction. 5 Schenkeveld (1975a) and Damon (1991) 45–49 discuss the role of ἄλογος αἴσθησις and the opposite notion of the ‘rational criterion’ ( λογικὸν κριτήριον ) in Dionysius’ critical works. The importance of ἄλογος αἴσθησις was also stressed by οἱ κριτικοί , who argued that the value of poetry was defined by the arrangement of the words, and that the latter aspect was judged upon by the irrational judgment: see Pohl (1968) 145–159 and Janko (2000) 120–189. Following Geigenmüller (1908) 34 and Nassal (1910) 37, Schenkeveld (1988) 304–305 convincingly argues that tacitus sensus at De or. 3.195 and Orat. 203 is Cicero’s translation of ἄλογος αἴσθησις . On the connection between Cicero’s ‘silent perception’ and Dionysius’ ‘irrational perception’, see section 4.3 n. 62. 6 See e.g. Dion. Hal. Comp. 15.12, Cic. Orat. 150 and ibid. 163. In section 4.3 below, we will see that rough arrangement was considered a disruption of the natural acoustic order.

148

OFFENDING THE EARS

‘arrangement types’ ( χαρακτῆρες τῆς συνθέσεως ) is based on this distinction: ‘smooth harmony’ ( ἁρμονία γλαφυρά ) consists largely of pleasing sounds, ‘rough harmony’ ( ἁρμονία αὐστηρά ) is dominated by grating sounds, and ‘well-mixed harmony’ ( ἁρμονία εὔκρατος ) is an appropriate combination of the two opposite effects (section 3.2.1).7 In this chapter, we will focus on the aesthetics of rough word arrangement, that is, on harsh, ear-jarring compositions that produce vexing rhythms and dissonant successions of tones. The significance of roughness in the Greek and Latin stylistic discourse can easily be overlooked: scholars usually discuss the acoustic effects of word arrangement under such headings as ‘euphony’ or the ‘euphonist tradition’.8 Yet, for a fuller understanding of the ancient views of word arrangement, we should also pay close attention to roughness and cacophony: if musicality is so important for the critics and rhetoricians in Rome, why do they concern themselves so extensively with disharmonious sounds that offend their ears? To be sure, roughness was not considered inherently vicious: when critics describe the acoustics of a passage as jarring or painful, they are not necessarily condemning it. Indeed, we already saw in the previous chapter that Dionysius associate rough word arrangement with such virtues as ‘beauty’ ( κάλλος ) and ‘grandeur’ ( μεγαλοπρέπεια ), and we will see below that his Greek and Roman colleagues also appreciate roughness in various ways and to various extents. 9 What is more, we will see that roughness played a major role in the lively debate about classical style in Rome: the production of harsh, grating sounds was sometimes seen as a means to echo, almost in a literal sense, the crude acoustics of centuries-old prose or poetry (section 4.5.3). Thus, rough composition could harness the sound of the Greek classics for the audiences in Rome. Yet, the opinions about the appropriate application of this procedure varied. It is the purpose of this chapter to outline the shared discourse on rough word arrangement and to shed

7 Dion. Hal. Comp. 21–24, Dem. 37–41. There appear to be close connections between Dionysius’ three types of arrangement and similar systems attributed to musicians (in Demetr. Eloc. 176) and to οἱ κριτικοί (in Philod. Poem. 5): in the latter two cases, a distinction is made between the adjectives ‘smooth’ ( λεῖος ), ‘rough’ ( τραχύς ) and ‘solid’, or ‘well-proportioned’ (εὐπαγής ). See esp. Pohl (1968) 99–100, 149–153. 8 Cf. e.g. Janko (2000) 165–189 and Porter (2016) 239–245. Of course, classicists are not unaware of the ancient appreciation of rough word arrangement. See e.g. Stanford (1967) 63–64: ‘A language, like an orchestra, needs clashing, clanging, and thundering instruments as well as a lighter wood-wind group; and a versatile author will need cacophonous sounds at times to express the harsher aspects of what he wants to say.’ Yet, as far as I am aware, the only study to date that focuses on the positive evaluation of roughness in ancient theories of word arrangement is Rispoli (1998): cf. the present section below. 9 See esp. section 3.2.1, table 6 above. On the connection between grandeur and roughness, see also section 4.5 below.

149

CHAPTER FOUR light on the goals and motivations that underlie the diverse evaluations of harsh and clashing combinations of words. In exploring the Greek and Latin discourse on rough word arrangement, we can build on several modern studies that touch on the ancient art of ‘composition’. Much work, for instance, has been done on ancient theories of prose rhythm and on ancient views about periodic sentence structure. 10 Aldo Scaglione’s monograph The Classical Theory of Composition will be of limited use for my purposes, as it exclusively tackles the ‘genuinely syntactical aspects of style’, while glossing over the musical-aesthetic aspects, which in the author’s view ‘have been studied extensively and rather satisfactorily’. 11 Ironically, however, the aural evaluation of word arrangement went on to receive considerable scholarly attention in the decades following the publication of Scaglione’s book. From the frazzled remains of Philodemus’ On Poems , for example, various classicists have been able to reconstruct the views of a group of Hellenistic critics, referred to by Philodemus as οἱ κριτικοί , who held that the value of poetry consisted primarily in its acoustic qualities. 12 In addition, scholars have

10 A good starting point for Latin theories of prose rhythm is Schmid (1959), who connects Cicero’s discussions about the topic to Arist. Rh. 3.8–9. As for Greek theory, one may start from Dover (1997), who offers a scathing review of the discussion in Dion. Hal. Comp. Gentili (1990) shows that Dionysius makes the abstract rules of meter and rhythm concrete by connecting them to the principle of irrational perception (‘instintiva percezione aurale’). Hutchinson (2013) 233–240 notes that the polemic about prose rhythm among Roman orators in Cicero’s day is related to ‘the Hellenistic system of rhythm, started, it is said, by Hegesias’ and to the practice of Greek declaimers in Rome. Concerning sentence structure, Innes (1994) has focused on Dionysius’ and Longinus’ conceptions of the period and the colon, while Viljamaa (2003) has addressed Dionysius’ use of the term colon and the comma. For good discussions of Dionysius’ use of linguistic and grammatical theories, see Schenkeveld (1983) and esp. De Jonge (2008). For Latin views on sentence structure, see Lieberg (1956) on the use of the term ‘structure’ (structura ). 11 Scaglione (1972) 24. The author traces the history of the theory of word arrangement from Antiquity to the present and, as a consequence, he is mostly interested in those aspects of the ancient discourse that have influenced later English, French and Italian theories, not necessarily the aspects which were considered paramount by the ancients themselves. This approach sometimes causes him to distort ancient notions by conflating them with later ones: thus, Aristotle’s views on the period are fitted into those of Quintilian and even equated with modern concepts. See the reviews of Winterbottom (1974) and Schenkeveld (1975b). 12 Schenkeveld (1968) was the first to discuss Philodemus’ οἱ κριτικοί as a unified school of aesthetic thought, which focused on euphony and word arrangement. Porter (1995) reevaluates the material and shows that Crates of Mallos was the main source of Philodemus’ information about οἱ κριτικοί : he concludes that Crates shared their obsession with the acoustic qualities of poetry, although the critic considered them his opponents. Janko (2000) 120–189 discusses the meaning of the term οἱ κριτικοί and links the various theses associated with them to individual authors, such as Heracleodorus and Pausimachus of Miletus, whose names survive only in Philod.

150

OFFENDING THE EARS been successful in uncovering the connections between musical theory and literary criticism: especially the discussions of melody and rhythm in the fragments of the Elements of Harmony and the Elements of Rhythm by the Peripatetic philosopher Aristoxenus of Tarentum (born ca. 370–365 BC) touch on several themes and topics that went on to become central to the discourse on word arrangement: both Dionysius and Cicero were familiar with his musicological work. 13 One theme that permeates the views of Aristoxenus, οἱ κριτικοί as well as the students of word arrangement in Rome is the focus on sensory perception in the critical evaluation of music, poetry and prose. James Porter has coined the helpful phrase ‘aesthetic materialism’, which refers to a focus on the audible, visible and tactile aspects of art: by shifting our attention from the dominant formalism and idealism of Plato and Aristotle to the ‘counter- tradition’ of sensualism and materialism, Porter has been able to demonstrate the centrality of perception and euphony throughout the history of Greek aesthetic thought. 14 Another recurring feature in ancient discussions about the acoustic effects of word arrangement is their imitative aspect, which scholars have subsumed under various categories, such as ‘expressiveness’, ‘onomatopoeia’, ‘mimesis’ and ‘iconism’. 15 Like ancient musicians, critics

Poem . See also Porter’s works listed in n. 14 below, which discuss the views of οἱ κριτικοί in the context of the history of ancient aesthetic thought. 13 Both Aristox. Harm. and Rhythm. survive incompletely; for the former we have the edition of Da Rios (1954), for the latter the editions of Pearson (1990) and Marchetti (2009). The extant data about the life and works of Aristoxenus are presented by Marchetti (2009) 1–25. Dionysius ( Comp. 14.2, Dem. 48.2) and Cicero ( Tusc. 1.19, 1.24) both refer twice to Aristoxenus. According to Kroll (1907) 91–101, Dionysius knew the musical theories of Aristoxenus through the work of Theophrastus. Koller (1954) 174–179, 193–202 traced the musical foundations of the theory of word arrangement further back to Democritus and the Pythagoreans; cf. Porter (1986) on the importance of atomism. Pohl (1968) links Dionysius’ theory of three harmonies to the musical doctrines of Aristoxenus and to the euphonic views of οἱ κριτικοί . The latter group is central in the discussion of Janko (2000) 173–189, who also summarizes the scholarly findings up to the end of the previous millennium. Recently, Rocconi (2010) has discussed four points of contact between Aristoxenus and Dionysius, including their focus on σύνθεσις and their attention to the relation between musical and conversational melody: cf. Barker (2014). 14 Porter (1986) presents materialist aesthetics as ‘a distinct counter-tradition, or else tendency, which occasionally breaks through the surface of the canonical, idealist and formalist tradition of Plato and Aristotle’, and which reaches its fullest known articulation in the lost works of οἱ κριτικοί , Dion. Hal. Comp. and Vitr. De arch. With this in mind, Porter (2010) revisits the history of ancient aesthetics, and Porter (2016) reevaluates the history of the sublime. Cf. also Porter (2001) on the sublime in Cicero, οἱ κριτικοί , Dionysius and Longinus. 15 A seminal study of mimesis in rhetorical theory and poetic criticism is Koller (1954), who stressed the influence of the Pythagorean idea that song, music and dance could cleanse the soul through the expression of

151

CHAPTER FOUR and rhetoricians (especially among the Greeks) advocated a correspondence between sound and substance: tone and rhythm should imitate and hence reinforce the meaning of the words (section 4.5.2). All in all, the art of word arrangement is a multifaceted subject, as Karin Pohl and Casper de Jonge have emphasized in their studies of Dionysius’ monograph on the topic: like many of his colleagues, the critic ‘incorporates views from all ancient language disciplines that are relevant to the subject’. 16 Lastly, I should mention the article by Gioia Rispoli on ‘roughness of sound’ ( δυσφωνία ), which, like the present chapter, focuses on the positive evaluation of harshness and cacophony. 17 Thus, the present study touches on various topics that have already been discussed in previous scholarship. Nonetheless, focusing on roughness and dissonance, it aims to contribute in two important ways to our understanding of the ancient theory of word arrangement. First, while classicists have been mostly interested in identifying the origins of the various doctrines that contribute to the discipline of word arrangement, I will adopt a synchronic perspective: I will not be concerned, therefore, with identifying the various sources of the critics and rhetoricians of Late-Republican and Augustan Rome, but I will rather focus on the programs, preferences and ambitions that govern their discussions. Secondly, this chapter will explore the relationship between Greek and Latin views on word arrangement. As we will see, Roman rhetoricians were acutely aware of the Greek

‘ethos’ and ‘pathos’ . In addition, Pohl (1968) 69–126 has connected several elements from Dion. Hal. Comp. to a ‘Mimesis-Lehre’ that she attributes to Peripatetic musical theory, esp. Aristoxenus. Wilkinson (1963) 46–88 discusses the ‘expressiveness’ of words in Latin. See also Asmis (2004), who argues that οἱ κριτικοί did not separate euphony and meaning, as is usually thought, but that they in fact insisted that sound served to emphasize substance. Calcante (2005a) presents a multidisciplinary approach to ‘iconism’, studying sources on music, poetry, prose, philosophy of language and the psychology of perception. In line with the previous study, Calcante (2005b) draws up a ‘system of euphony’, based predominantly on his readings of Dion. Hal. Comp. and Pl. Cra. (mentioned at Comp. 16.4). 16 De Jonge (2008) 42–43. De Jonge rightly emphasizes Dionysius’ pragmatic approach to his sources: ‘We find Dionysius selecting the workable ideas from different language sciences, while at the same time avoiding elaborate discussions of technical details that are not useful for his intended audience.’ Pohl (1968) offers a rich study into Dionysius’ system of three types of arrangement: she distinguishes between three groups of sources, namely musical, rhetorical and poetical theory. Due to this interdisciplinary application, it is all but impossible to assign the individual notions associated with word arrangement to any single source or tradition: cf. the discussion about Dionysius’ sources for his chapter on rhythm ( Comp. 17) in De Jonge (2008) 341–342 n. 60. 17 Rispoli (1998) takes Philod. Poem. as her starting point and notes that there are striking parallels with Demetr. Eloc. and Dion. Hal. Comp. , which she traces back to Greek musical doctrines on the connections between music and speech. Unlike Rispoli’s diachronic approach, this chapter adopts a synchronic perspective to composition.

152

OFFENDING THE EARS achievements in the discipline (section 4.2), and they applied similar theories, techniques and terminology as their Greek colleagues (sections 4.3 and 4.4). Yet, there is notable variation among the extant evaluations of rough word arrangement (section 4.5): these differences can sometimes be attributed to ad hoc writing purposes and individual tastes, but we will also see that Roman ears did not appreciate roughness in the same way as Greek ears.

4.2 Word Arrangement: a Greek Discipline in Rome There is a major difference between Greek and Roman attitudes to rough word arrangement, which goes back to the second century BC: it is, thus, as old as our record of word arrangement in Rome. In 168 BC, the Pergamese grammarian and critic Crates of Mallos is reported to have visited the city as an envoy of the Attalid court: during his stay, he allegedly fell into a sewage hole near the Palatine, broke his leg and spent his recovery giving lectures about grammar and literary criticism. 18 According to Suetonius, Crates, whose many interests included word arrangement and euphony, inspired Romans to read, emend and annotate Latin poetry, initially only the poems written by their friends, but later also the verses of Naevius, Ennius and Lucilius. 19 I have included this quaint anecdote not in order to prove that there was no study of euphony or word arrangement in Rome before Crates’ visit, but rather to illustrate the widely shared sentiment among Romans that they inherited their literary wisdom from the Greeks. 20 In this section, we will see that Roman scholars stressed their indebtedness to their Greek predecessors, all the more emphatically, in discussions of word arrangement. In fact, various Latin sources present the careful arrangement of words as a quintessentially Greek activity, to the extent that it might even be awkward for Romans to create smooth, euphonious collocations.

18 The story about Crates’ visit to Rome can be found in Suet. Gramm. 2.1–4, who dates the embassy shortly after Ennius’ demise (169 BC): on the textual problems of the passage, see Vacher (1993) 40 n. 3. According to Suetonius, Crates ‘was the first to introduce the study of grammar to the city’ ( primus studium grammaticae in urbem instituit ). Lehmann (2004) focuses on the passage as a source for the beginning of criticism in Rome. 19 Suetonius refers to three philological activities that the Romans learnt from Crates, i.e. ‘reading’ ( legere ), ‘emending’ ( retractare ) and making ‘annotations’ ( commentarii ). See Lehmann (2004) 151–162, who explores the Roman critical activities with respect to Naevius, Ennius and Lucilius in the second century BC. For Crates’ intellectual accomplishments, see Pfeiffer (1968) 235–238. For Crates’ critical views about arrangement and euphony, and his role in Phil. Poem. , see Janko (2000) 120–134. 20 Cf. section 3.3 above on the Athenian embassy of 155 BC and the introduction of the three styles in Rome. Blaensdorf (1994) 8–10 has shown that already in the early second century BC (well before Crates’ visit) the Romans adopted critical approaches to legal and religious texts.

153

CHAPTER FOUR

The fragments of the satirist Lucilius are the earliest extant Latin sources on word arrangement. The poet was well-versed in the Greek terminology of acoustic criticism: he classified sonorous words as ‘euphonious’ ( εὔφωνα ) and he submitted that individual letters could be ‘combined in an ugly way’ ( κακοσύνθετα ). 21 We also have a couplet of hexameters that mock the word arrangement of the orator and infamous hellenomaniac T. Albucius. In 120 BC, Albucius filed an unsuccessful lawsuit against the praetor Q. Mucius Scaevola, in whose mouth Lucilius, in his second book of Satires , placed the following insult at the expense of Albucius: 22

Quam lepide λέξεις compostae ut tesserulae omnes arte pavimento atque emblemate vermiculato.

How pleasantly les mots are joined together, like little cubes all artfully inlaid in a worm-like mosaic floor.

The point of these verses is that Albucius shows himself a true Greek (which is not a good thing in Scaevola’s book) by indulging in a fundamentally Greek activity: he arranges his words neatly into an intricate, overelaborate design. 23 Architectural imagery is common in ancient discussions of word arrangement: while overly smooth combinations are like the

21 See Janko (2000) 176: Lucilius ‘knew a complete theory of σύνθεσις ’. Lucil. fr. 1188 Krenkel (= fr. 418 Warmington): the text refers to ‘words that are more sonorous’ ( quae verba magis sonantia sunt ), and ‘which Lucilius calls euphonious’ ( quae Lucilius εὔφωνα appellat ). Lucil. fr. 367–368 Krenkel (= fr. 389–390 Warmington): the letter r is inherently ugly, even if it is not ‘combined in an ugly way’ ( κακοσύνθετον ) or pronounced ‘in dog-language’ ( canina lingua ). The word κακοσύνθετος is frequent in scholia, e.g. on Eur. Or. 674, Hec. 801, Ar. Vesp. 818. For Lucilius’ statements on euphony, cf. also Lucil. fr. 366 Krenkel (on the letter q), fr. 369–370 Krenkel (on the letter s). 22 Lucil. fr. 74–75 Krenkel (= fr. 84–85 Warmington), my translation; French words to express Albucius’ affectation are standard in English translations, cf. Warmington (1938) 29. For the altercation between Albucius and Scaevola, cf. Gruen (1992) 257–258, 290–291 on Cic. De or. 2.281, 3.171; Brut. 102, 113; Fin. 1.8. 23 For this point, see Chahoud (2004) 31–37, who shows that Lucilius’ incorporation of Greek (code-switching) serves to underline an important social rule: ‘The Romans must behave like Romans and speak like Romans, unless they aim to be laughed at, and deservingly so.’ Butler (2011) 39–42 focuses on the ‘worm-like’ (vermiculatus ) appearance of the mosaic (cf. the modern architectural term opus vermiculatum ): ‘From close by, the array of wriggling lines can indeed (with a little imagination) resemble a bed of worms.’ Yet, ‘any viewer need only move closer and the illusion dissolves’: thus, under careful scrutiny, the ‘worm-like’ style of verbal arrangement falls short. Lucilius imitates Albucius’ smooth arrangement through a series of four elisions.

154

OFFENDING THE EARS cubes in a mosaic floor, rough combinations can be compared to the crudely cut building blocks of a rugged structure. In Dionysius’ words, rough harmony ‘does not mind admitting harsh and dissonant collocations, like blocks of natural stone laid together in building, with their sides not cut square or polished smooth, but remaining unworked and rough-hewn’. 24 Lucilius does not refer to such an uneven, jagged edifice, but rather to a sleek, well-fitting structure: he presents the view that an immaculately smooth arrangement is unbefitting a true Roman. His satirical objections are cited with approval by Cicero, who condemns Albucius’ labor as ‘both endless and silly’ ( cum infinitus tum puerilis ), and by Quintilian, who compares the philhellene orator to a rider who kills the passion of his fiery horse by forcing it to perform a dainty dance routine. 25 The idea that neatly ordered words are typical for Greek prose practice persists in Latin texts of the first centuries BC and AD. The Roman rhetoricians do not even exhibit a coherent Latin terminology for the topic of arrangement: Rhetorica ad Herennium and Quintilian use compositio , while Cicero uses collocatio in his Orator , although he had simply subsumed the topic under the category of ‘periodic structure’ ( continuatio verborum ) in De Oratore. To complicate matters even further, in Cicero’s discussions the terms collocatio and compositio sometimes denote a subtopic (and not the whole subject) of word arrangement, namely the combining of final syllables with the subsequent initial syllables, for which Quintilian uses iunctura. 26 On several occasions, Cicero presents Latin technical terms

24 Dion. Hal. Comp. 22.2: Τραχείαις τε χρῆσθαι πολλαχῇ καὶ ἀνιτύποις ταῖς συμβολαῖς οὐδὲν αὐτῇ διαφέρει , οἷαι γίνονται τῶν λογάδην συντιθεμένων ἐν οἰκοδομίαις λίθων αἱ μὴ εὐγώνιοι καὶ μὴ συνεξεσμέναι βάσεις , ἀργαὶ δέ τινες καὶ αὐτοσχέδιοι . This passage is discussed at length by De Jonge (2008) 204–213, with respect to Dionysius’ grammatical theory of the parts of speech. For the comparison between word arrangement and architecture in Greek and Latin stylistic theory, see De Jonge (2008) 186–204, who adduces examples from Demetr., Philod., Dion. Hal., Long., Cic. and Quint. Cf. esp. Demetr. Eloc. 13, where the periodic style is compared to a well-fitted edifice, while ‘the clauses of the disjointed style (resemble) stones which are simply thrown about near one another and not built into a structure’ ( τὰ δὲ τῆς διαλελυμένης ἑρμηνείας διερριμμένοις πλησίον λίθοις μόνον καὶ οὐ συγκειμένοις). De Jonge (2016) 64–65 points out similarities between Dion. Hal. and Vitruv. De arch. (e.g. 1.2.2: apta conlocatio ). 25 Cic. De or. 3.171, Brut. 274, Orat. 149. Quint. Inst. orat. 9.4.113: ‘Will not his passion cool and his energy flag, just as showy riders spoil the free movement of their horses by forcing them to a mincing gait?’ ( nonne ergo refrigeretur sic calor et impetus pereat, ut equorum cursum delicati minutis passibus frangunt? ). MacPhail (2014) 92–115 devotes a chapter to the comparison between mosaic and text in Antiquity (Lucil., Cic., Quint.) and in Renaissance literature. Plin. NH 36.185 also quotes Lucilius’ verse in a his discussion of mosaic styles. 26 Compositio and collocatio are obvious translations of the Greek word σύνθεσις : they denote the entire topic of word arrangement at e.g. Rhet. Her. 4.18, Cic. Orat. 234, Quint. Inst. orat. 9.4.1. The word compositio is used by

155

CHAPTER FOUR explicitly as translations from the Greek: he discusses, for instance, ‘that rounded form of expression, which the Greeks call the περίοδος , and to which we apply the term ambitus , comprehensio , continuatio or circumscriptio ’. Similarly, he refers to the phrases that make up the period as ‘those forms which, as the Greeks call them κόμματα and κῶλα , we might properly call incisa and membra ’. In addition, he does not introduce the Latin word for rhythm ( numerus ) without mentioning the Greek word ( ῥυθμός ) as well: the Latin probably reflects Aristotle’s association of rhythm with ‘number’ ( ἀριθμός ). 27 Such explicit translations can still be found in Quintilian’s discussion of word arrangement: 28 thus, despite their own sizeable achievements on the topic, the Romans continued to emphasize its Greekness. As a result of this general attitude, Roman rhetoricians often display a distinct unease, when they launch into a discussion about the intricacies of word arrangement. The short passage about the topic in Rhetorica ad Herennium , for instance, is limited to a list of six don’ts: the orator is advised to avoid frequent hiatus, alliteration, polyptoton, homoeoptoton, hyperbaton and long periods. 29 In Cicero’s De oratore , the protagonist Crassus is constantly on his guard, lest he be denounced as ‘some idle and jabbering little Greek’ ( aliqui Graeculus

Cic. Orat. 201 to refer to the first of three tasks of word arrangement, namely to make sure ‘that final syllables may fit the following initial syllables as neatly as possible’ ( ut inter se quam aptissime cohaereant extrema cum primis eaque sint quam suavissimis vocibus ): this same topic is called collocatio in Cic. De or. 3.171, and iunctura in Quint. Inst. orat. 9.4.32. For continuatio (the heading of the section on word arrangement in Cic. De or. 3.171) as Cicero’s translation of περίοδος , see n. 27 below. Cf. also section 4.4 n. 63–65 below for the Greek and Latin divisions of the subject of word arrangement. 27 Cic. Orat. 204: In toto circuitu illo orationis, quem Graeci περίοδον , nos tum ambitum, tum circuitum, tum comprehensionem aut continuationem aut circumscriptionem dicimus. Ibid. 211: Illa quae nescio cur, cum Graeci κόμματα et κῶλα nominent, nos non recte incisa et membra dicimus. Ibid. 170: Numerus Latine, Graece ῥυθμός . Cf. Arist. Rh. 3.8.1–3: ‘Now all things are limited by number, and the number belonging to the form of diction is rhythm’ ( περαίνεται δὲ ἀριθμῷ πάντα· ὁ δὲ τοῦ σχήματος τῆς λέξεως ἀριθμὸς ῥυθμός ἐστιν ). Transl. Kennedy (1991). According to Formarier (2013), the word numerus allows Cicero to stress the huge variety of the possible rhythmical combinations in prose. For other explicit Latin translations of Greek terms, cf. Orat. 134 (χαρακτήρ , forma ipsa ), ibid. 166 ( ἀντίθετα , contraria ) and ibid. 181 ( σχήματα , quasi formae et lumina ). 28 Quint. Inst. orat. 9.4.22 ( κόμμα , incisum ; κῶλον , membrum ; περίοδος , ambitus, circumductus, continuatio, conclusio ), 9.4.36 ( συναλιφαί , coeuntes litterae ) and 9.4.45 ( ῥυθμός , numerus ; μέτρον , dimensio quaedam ). 29 Rhet. Her. 4.18. The author does not use the Greek or Latin technical terminology, but he rather gives descriptions, e.g., ‘excessive recurrence of the same letter’ ( eiusdem litterae nimia adsiduitas ) for alliteration, ‘excessive repetition of the same word’ ( eiusdem verbi adsiduitas nimia ) for polyptoton, and ‘series of words with like case endings’ ( similter cadentia verba ) for homoeoptoton. Cf. ibid. 4.10 and Calboli (1959) 305–306. The author illustrates his discussion with citations, some of which can be assigned to Ennius, e.g. o Tite, tute, Tati, tibi tanta, tyranne, tulisti. On (the absence of) hiatus in Latin according to Cicero, see section 4.5.1 below.

156

OFFENDING THE EARS otiosus et loquax ): 30 he commends Lucilius’ verses for their humor, but he adds that ‘it is nevertheless important to pay attention to this matter of verbal arrangement’. 31 In his Orator , Cicero attaches a long preface to his discussion of word arrangement, which serves as an apology for his engagement with this high-brow, technical activity: ‘I am going to speak about the arrangement of words and almost about the counting and measuring of syllables.’ He presents the theoretical nature of this exercise as unusual for a Roman: ‘I am inclined to think that most Roman orators had more talent than instruction ( doctrina ); consequently they were better able to speak than to lay down precepts ( praecipere ), but with us perhaps just the contrary is true.’ 32 The activity of teaching was, of course, seen as a Greek specialty, but as Cicero submits, it is the only thing left to do for a retired orator and scholar like himself. 33 Cicero is aware, then, that devoting half of his Orator to the intricacies of word arrangement might be suspect. Yet, he makes an elaborate case for its study, claiming that it helps create a style that is ‘smooth’ ( levis ), ‘pleasant’ ( suavis ) and appealing to the ‘pleasure of the ears’ ( voluptas aurium ). 34 By doing so, Cicero defends his own oratorical legacy against the attacks of his opponents. Specifically, his oratory was censured for its obtrusive

30 Cic. De or. 1.102. See Wisse (2002) 334–341 on the relation between the portrayal of Greek learning in De or. and the contemporary spectrum of Roman attitudes to Greek culture: ‘It was clearly important to Cicero at this point to counter the possible impression that Crassus’ mastery of Greek intellectual subjects made him “un- Roman”, and to leave no doubt that admiration of Greek intellectual accomplishments did not stand in the way of healthy feelings of Roman superiority.’ 31 Cic. De or. 3.172: Sed est tamen haec conlocatio conservanda verborum . Transl. May and Wisse (2001). The word conlocatio refers exclusively to the activity of combining final syllables with subsequent initial syllables: cf. n. 26 above. Crassus not only quotes Scaevola’s remarks about Albucius’ mosaic-like arrangement (Lucil. fr. 74–75 Krenkel = fr. 84–85 Warmington), but also a reference by Scaevola to Crassus himself (Lucil. fr. 76 Krenkel = fr. 86 Warmington): ‘I have Crassus as a son-in-law, so don’t be too much of an orateur ’ ( Crassum habeo generum, ne ῥητορικώτερος sis ). At De or. 171, Crassus takes this to mean that his own style of word arrangement is to be preferred over the Albucius’ philhellenic style. 32 Cic. Orat. 140–148. See esp. 147: De verbis enim componendis et de syllabis propemodum dinumerandis et demetiendis loquemur. Ibid. 143: Atque haud scio an plerique nostrorum oratorum ingenio plus valuerint quam doctrina; itaque illi dicere melius quam praecipere, nos contra fortasse possumus. 33 Cicero refers to his lifelong engagement with literary studies ( Orat. 146) and his current otium (ibid. 148) as excuses for taking up the technical study of word arrangement. Dugan (2005) 253–267 relates Cicero’s apology at Orat. 140–148 to an intellectual context (the abolition of the tirocinium fori ), a personal context (his friendship with Brutus) and a political context (the dictatorship of Caesar). I do not agree with Dugan that Cicero’s turn to private teaching springs from his resignation to the loss of the Republic: cf. section 5.5.1. 34 For Cicero’s emphasis on aural satisfaction ( voluptas aurium ), see e.g. De or. 3.177, 3.180; Orat. 159, 198, 203, 208, 237. Cf. section 4.4 below on the Latin terminology for rough and smooth arrangement.

157

CHAPTER FOUR and affected rhythms by Calvus and the so-called ‘Atticists’: ‘It seems too much like a trick to catch the ear, if the orator in the midst of his speech is hunting for rhythms. Relying on this objection, those terrible men themselves deliver broken and choppy sentences and upbraid those who produce rounded and finished periods.’ 35 Cicero’s focus on smooth composition remained controversial after his death: the phrase esse videatur (i.e., a paean followed by a spondee) became a symbol for Cicero’s excessive fondness for rhythmical cadences. 36 Even Quintilian criticizes his hero on this issue, expressing his own predilection for roughness over smoothness: ‘In general, however, if I had to choose, I should prefer the arrangement to be hard and harsh rather than effeminate and emasculated, such as the kind we see in many writers (and more and more day by day), that dances to the lascivious tunes of the castanet.’ 37 Thus, we may conclude that among Roman authors in Late-Republican and Early- Imperial Rome rough, unwrought collocations generally landed on sympathetic ears, while smooth, carefully arranged words were mistrusted: for this reason, Cicero’s project advocating a meticulous study of euphony and word order required a lengthy preamble. Greek critics, conversely, could make more confident statements about the importance of arrangement. According to Dionysius, for instance, ‘this is what makes the essential difference between one poet or orator and another—the dexterity with which they arranged

35 Cic. Orat. 170: Nimis enim insidiarum ad capiendas auris adhiberi videtur, si etiam in dicendo numeri ab oratore quaeruntur. Hoc freti isti et ipsi infracta et amputata locuntur et eos vituperant qui apta et finita pronuntiant. For the views of Calvus and the Atticists, see section 3.4 above and esp. sections 5.2 and 5.6 below. 36 The phrase is criticized by Quint. Inst. orat. 9.4.73, 9.4.145 (the conclusion of the ninth book, which closes on the words esse videantur , a playful jibe at Cicero) and 10.2.18. Tac. Dial. 23.1 has Aper refer to the phrase as the ‘tag which he tacks on as a meaningless catchphrase in every second sentence throughout his speeches’ ( illud tertio quoque sensu in omnibus orationibus pro sententia positum ). Zieli ński (1904) lists the rhythmical pattern of esse videatur (― ᴗ ᴗ ᴗ ― ―) as a variation of Cicero’s first ‘Hauptform’, i.e., ― ᴗ ― ― x with the second long syllable resolved. It seems, however, that neither esse videatur nor phrases with the same pattern (e.g., posse videamur ) was unduly common in Cicero’s speeches: Powell (2013) 59–65. The use of meaningless patch- words was associated with (overly) smooth word arrangement in ancient stylistic theory: see Sluiter (1997) 238– 244 with e.g. Cic. Orat. 230, Demetr. Eloc. 55–58, and Dion. Hal. Comp. 22.5. 37 Quint. Inst. orat. 9.4.142: In universum autem, si sit necesse, duram potius atque asperam compositionem malim esse quam effeminatam et enervem, qualis apud multos, et cotidie magis, lascivissimis syntonorum modis saltat. Quintilian hesitantly announces his deviation from Cicero at the beginning of his chapter on word arrangement ( Inst. orat. 9.4.1): ‘On some points I may express a slightly different opinion’ ( in quibusdam paulum fortasse dissentiam ). Formarier (2013) shows that Quint. Inst. orat. 9.4.53–55, unlike Cic. Orat. , avoids the charge of confusing oratory with music by distinguishing explicitly between oratorical oratorical rhythm (which he calls numerus ) and musical rhythm (which he calls rhythmos ).

158

OFFENDING THE EARS their words’. 38 Similar views are attributed to οἱ κριτικοί by Philodemus: ‘The best poets rise to the first rank and they alone endure on no other account than the sounds.’ 39 In Demetrius’ On Style , acoustic word arrangement is arguably presented as the most important aspect of prose style: concerning Thucydides, for instance, the author says that ‘while he has the full range of grandeur, it is perhaps this power of verbal arrangement which alone or chiefly secures his greatest grandeur’. 40 Longinus’ On the Sublime , lastly, lists composition as the last, but not the least, source of the sublime, ‘affecting not only the ears but the very soul’ and, if used successfully, ‘winning a complete mastery over our minds’. 41 All of this does not mean that there was a complete agreement among Greek critics about the importance of careful word arrangement. Philodemus, for one, deprecates the attention paid to euphony and arrangement by οἱ κριτικοί , claiming that poetry and prose should please the intellect rather than the ear. 42 In On the Arrangement of Words and On

38 Dion. Hal. Comp. 4.13: Καὶ τοῦτ’ ἦν σχεδὸν ᾧ μάλιστα διαλλάττει ποιητής τε ποιητοῦ καὶ ῥήτωρ ῥήτορος , τὸ συντιθέναι δεξιῶς τὰ ὀνόματα . Dionysius compares the function of word arrangement in style to the role that the goddess Athena plays in Hom. Od. , having the ability to change the appearance of Odysseus in whatever way she wishes: see section 2.4.3 n. 145 and section 3.5 n. 138 above. 39 Philod. Poem. 1 col. 83.11–14 Janko: Οἱ ἀγαθοὶ ποηταὶ παρ ’ οὐδὲν ἄλλο πρωτεύουσίν τε καὶ μόνοι διαμένουσιν ἢ παρὰ τοὺς ἤχους . The text refers to euphony, not to arrangement, which may be relevant, as the two topics were sometimes carefully distinguished in Hellenistic poetic criticism. See e.g. Philod. Poem. 5 col. 24.27–31 Mangoni: ‘Crates misunderstands the views of Heracleodorus and those who share them; for they praise not the arrangement, but the sound which supervenes upon it’ ( ἀποτυγχάνει τοιγαροῦν τῆς Ἡρακλειοδώρου καὶ τῶν ὁμοίων δόξης· οὐ γὰρ τὴν σύνθεσιν , ἀλλὰ τὴν ἐπιφαινομένην αὐτῇ φωνὴν ἐπαινοῦσι ). This distinction between euphony and arrangement is particular to Hellenistic criticism and is, thus, not relevant for our understanding of the sources discussed in this dissertation: see Porter (2016) 241 n. 161. 40 Demetr. Eloc. 40: Κινδυνεύει τῷ ἀνδρὶ τούτῷ παντοδαποῦ ὄντος τοῦ μεγαλπρεποῦς αὕτη ἡ σύνθεσις μόνη ἢ μάλιστα περιποιεῖν τὸ μέγιστον . Word arrangement is discussed on numerous occasions throughout the treatise (cf. section 4.1 n. 4 above): the author not only describes the characteristics of word arrangement in each of the four styles, but he also opens his treatise with a long preface on periodic sentence structure (ibid. 1–35). 41 Long. Subl. 39.3: Τῆς ψυχῆς αὐτῆς, οὐχὶ τῆς ἀκοῆς μόνης ἐφαπτομένων . Ibid.: Παντοίως ἡμῶν τῆς διανοίας ἐπικρατοῦσαν . At Subl. 8, the author introduces arrangement as the activity, ‘which gives form to all sources already mentioned’ ( συγκλείουσα τὰ πρὸ ἑαυτῆς ἅπαντα ). On the magical powers of word arrangement according to Longinus (and Dionysius), see De Jonge (2008) 332–340 and De Jonge (2012a) 291–292. 42 Positive articulations of Philodemus’ own views in Poem. are scarce: yet, Philod. Poem. 5 col. 23.26–24.11 Mangoni stresses the importance of ‘reason’ ( λόγος ) and ‘thought’ ( διάνοια ) in the evaluation of poetry. Philodemus’ objections to the emphasis on aural perception are not limited to poetic criticism: in Rhet. 4 col. 2.2–10 p. 163 Sudhaus, Philodemus attacks a group of ‘sophists’ ( σοφισταί ) who ‘reduce’ the matter of hiatus ‘to the pleasure and displeasure of the ear’ (πρὸς τὴν τῆς ἀκοῆς ἡδονὴν καὶ ἀηδίαν ἀναφέρουσιν ).

159

CHAPTER FOUR

Demosthenes , Dionysius feels the need to defend the careful attention that his favorite orator paid to the topic. One of the arguments that he makes in favor of Demosthenes refers to the necessity to look after the details, when discussing important matters: ‘It appears to me far more appropriate in a man who is composing political speeches which are to be permanent memorials to his powers, that he should not ignore even the smallest details, than it is for painters and engravers, who display their manual skills and industry upon perishable materials, to exhaust the refinements of their artistry on fine veins, young plumage, the first beard’s down and minute details of a similar character.’43 This passage responds to the early detractors of Demosthenes, such as Aeschines and Demetrius of Phalerum, who objected to the orator’s verbal trickery.44 Yet, Dionysius’ words also call to mind Cicero’s ‘counting and measuring of syllables’: Dionysius’ defense of Demosthenes’ carefully arranged words may serve to convince his skeptical Roman audience, in the first place his student Rufus Metilius, of the importance of the art of verbal composition.45 This section has pointed out a fundamental difference of approach between Latin and Greek discussions of word arrangement. For the Romans, the topic was generally considered a prerogative of Greek scholarship: as a result, the Roman orator whose tones and rhythms were perceived as smooth, could be denounced for spending too much time on an un-Roman, unmanly activity. Greek critics, such as Demetrius, Dionysius and Longinus, on the other hand, did not hesitate to ascribe tremendous powers to the art of arrangement, whether the words be ordered smoothly or roughly. As we will see below, this basic distinction between Greeks and Romans is reflected in their respective discussions about the stylistic applications

43 Dion. Hal. Comp. 25.29–44, Dem. 51.2–6. See esp. Comp. 25.35: Πολύ τε γὰρ μᾶλλον ἐμοὶ δοκεῖ προσήκειν ἀνδρὶ κατασκευάζοντι λόγους πολιτικοὺς μνημεῖα τῆς ἑαυτοῦ δυνάμεως αἰώνια μηδενὸς τῶν ἐλαχίστων ὀλιγωρεῖν, ἢ ζωγράφων τε καὶ τορευτῶν παισίν , ἐν ὕλῃ φθαρτῇ χειρῶν εὐστοχίας καὶ πόνους ὑποδεικνυμένοις , περὶ τὰ φλέβια καὶ τὰ πτίλα καὶ τὸν χνοῦν καὶ τὰς τοιαύτας μικρολογίας κατατρίβειν τῆς τέχνης τὴν ἀκρίβειαν . 44 Aeschines accused Demosthenes of deceiving his audience through the arrangement of his words (Aeschin. Ctes. 142) and by exploiting the tone of his voice (ibid. 210). Cf. Dion. Hal. Dem. 35 for additional criticisms in Aeschin. Fals. leg. 34, 40. According to Porter (2010) 319–322, ‘there seems to have existed already by this time a kind of theoretical reflection on the subject to which Dionysius of Halicarnassus, possibly in imitation, would later devote a treatise—using the same phrase as is found in Aeschines (and nowhere else earlier) for his title and in his writings.’ Indeed, Aeschin. Ctes. 142 has τὴν τῶν ὀνομάτων σύνθεσιν , cf. Anaximenes Rh. Al. 22.8, 25.1, 25.3–4. Other criticisms of Demosthenes’ focus on style can be found in Demetrius of Phalerum (fr. 162, 169 Wehrli), Hieronymus of Rhodes (fr. 52a Wehrli) and other sources, listed by Marchiori in Donadi and Marchiori (2013) 383 n. 34. Cf. section 2.3.1 above on the Hellenistic reception of Demosthenes. 45 Cic. Orat. 147. On the possibility that Dionysius reacts to his Roman coevals, see also section 4.5.1 below.

160

OFFENDING THE EARS of roughness (section 4.5). First, however, I will argue that despite their differences Greek and Roman critics still participated in a shared discourse on word arrangement.

4.3 The Universal Law of Arrangement and the Judgment of the Ear How do the ancient critics and rhetoricians distinguish between smooth and rough arrangements? Why, for instance, does Dionysius describe the opening line of Thucydides, beginning with the words Θουκυδίδης Ἀθηναῖος ξυνέγραψε , as ear-jarring? And why does Cicero insist that a sentence by C. Gracchus, ending on the words qui improbos probet , is harsh-sounding? 46 As we will see, roughness and smoothness were not regarded as matters of personal taste: Cicero, Dionysius and their colleagues subscribe to the thesis that there exists a universal law of nature for combining sounds. This law forms the backbone of the joint theoretical framework on which the extant Greek and Latin discussions of word arrangement are built: collocations that abide by the universal law are classified as pleasant and smooth, while those that transgress against it are considered vexing and rough. On the basis of their shared understanding of nature’s dictates, the authors generally agree on the sources for roughness: hence, as we will see later, the same basic rules for creating harsh, unpleasant compositions are applied to Greek and Latin literature alike (section 4.4). There are three essential aspects of the natural law of arrangement that we should take into consideration. These three components, or rules, do not merely recur frequently in rhetorical theory and literary criticism, but they are deeply ingrained in ancient thought about the properties of ‘articulate sound’ ( φωνή , vox ): as such, they are core features in many fields of knowledge, including music, rhythm, meter, grammar and the philosophy of language. 47

46 Dion. Hal. Comp. 22.37 on Thuc. 1.1.1; Cic. Orat. 233 on C. Gracchus fr. 48.24 Malcovati. The latter fragment is taken from Gracchus’ speech from 124 BC before the censors, defending himself against the charge of dereliction of duty by leaving his post as governor of Sardinia: cf. Plut. Gracch. 2–3. The full quote runs as follows: ‘It is inevitable that the man who approves of the wicked will disapprove of the good’ ( abesse not potest quin eiusdem hominis sit probos improbare qui improbos probet ). Cicero claims that the sentence would be ‘better fitted’ ( aptius ), if it had ended on the words qui improbos probet probos improbare. See the present section below. 47 On the interdisciplinary approach to word arrangement, see section 4.1 above. For the opposition between ‘articulate sound’ ( φωνή ) and ‘inarticulate sound’ ( ψόφος ), see Ax (1978). On the concept of φωνή in rhetoric and related disciplines, see e.g. Schenkeveld (1990), who outlines the structure and influence of Stoic theories on the voice; Porter (2010) 308–404, who discusses the role of the voice in materialist aesthetics; and Schulz (2014) 351–376, who exposes the connections between rhetorical theories on the voice (esp. concerning delivery, or ὑπόκρισις /actio ) and other sciences, such as grammar, drama, philosophy, music and medicine.

161

CHAPTER FOUR

1. The universal law of arrangement posits that in composing a piece of music, poem or prose text not just any sound-unit can be placed after just any other, but rather that there are certain unchangeable rules, ordained by ‘nature’ ( φύσις , natura ), that limit the number of possible combinations. 2. This law of nature is firmly entrenched in the human consciousness: all human beings possess the innate ability to distinguish between acoustic combinations that follow the natural law and those that break it. This ability does not depend on education, training or logical consideration, but rather on subconscious intuition: everyone can judge combinations of sound on the basis of their ‘irrational perception’ ( ἄλογος αἴσθησις , tacitus sensus ). 3. This instinctive faculty is equally strong in experts and in laymen: both groups can establish whether a given combination is in accordance with the universal law or not. Only experts, however, can explain why a given combination follows or violates the dictates of nature. They do so by breaking down the composition into its primary, indivisible ‘elements’ ( στοιχεῖα, elementa ) that form the building blocks of any articulate utterance in music, poetry and prose.

The search for elementary sound-units ( στοιχεῖα) is an inherent feature of ancient Greek thought about language and sound. 48 For most grammarians the primary elements are the individual letters ( γράμματα ), which can be combined into syllables and words. Rhythmicians and metricians, however, divide the composition into periods of time (χρόνοι ), which form feet and verses. In musical theory, lastly, the arrangements can be broken down into such elements as notes ( φθόγγοι ) and their intervals ( διαστήματα ), which build scales and chords. 49

48 Porter (2010) 213–239 shows that the concept of elements ( στοιχεῖα) not only plays an important role in theories about physics and metaphysics, but also in ancient analyses of music, poetry and prose: ‘Atoms, the building blocks of the universe (its stoicheia ), lie beneath the threshold of sensation; joined together in compounds ( suntheseis ) they produce visible phenomena: sounds, colors, smells, and so on.’ It follows that every human being can perceive and evaluate the compounds, but the study of their constituent elements is the exclusive prerogative of expert scholars. 49 See Koller (1955) for an overview of the different meanings of the term στοιχεῖον in Greek theories of language and music. In grammatical contexts, the words στοιχεῖον and γράμμα are often not clearly distinguished from each other, though sometimes they carry different meanings: cf. Sluiter (1990) 44 n. 19. Furthermore, note that the two partially surviving Aristoxenean treatises on music are called Elements of Harmony (Ἁρμονικὰ στοιχεῖα) and Elements of Rhythm (Ῥυθμικὰ στοιχεῖα): for the harmonic elements, see esp. Aristox. Harm. 27.22–32 (quoted below); for the rhythmic elements, see esp. Aristox. Rhythm. 2.11, Aristid.

162

OFFENDING THE EARS

Although the rough material is different across the various sciences, the natural law of arrangement does not change. In his Elements of Harmony, the musical expert Aristoxenus puts forward this point quite elegantly with respect to combining letters and musical notes: ‘In speaking, it is natural ( φύσει ) for the voice, in each syllable, to place some one of the letters first, others second, third and fourth, and so on for the other numbers. It does not place just any letter after any other: rather, there is a kind of natural growth in the process of putting together ( τοιαύτη τις φυσικὴ αὔξησις τῆς συνθέσεως ). In singing, similarly, when the voice places intervals and notes in succession, it appears to maintain a natural principle of combination ( φυσική τις σύνθεσις ), and not to sing every interval after every other.’ 50 Aristoxenus’ ‘natural principle of combination’ (i.e., rule 1 listed above) also applies to the art of word arrangement in poetry and prose. In critical and rhetorical theory, however, individual words are identified as the elementary building blocks: in On the Arrangement of Words , Dionysius aptly calls the parts of speech the ‘elements of discourse’ ( στοιχεῖα τῆς λέξεως ). 51 According to Dionysius, the putting together of words is quite similar to composing music: ‘The science of civic oratory is, after all, a kind of musical science (μουσική τις ), differing from vocal and instrumental music in degree, not in kind. In oratory as in music, the phrases possess melody, rhythm, variety and appropriateness; so that here too the ear delights in the melodies, is stirred by the rhythms, welcomes the variations, and all the time desires what is appropriate to the occasion. The distinction is simply one of degree.’ 52

Quint. Inst. orat. 1.14. Vollgraff (1949) discusses the early attestations of the Latin word elementum as a translation of στοιχεῖον . 50 Aristox. Harm. 27.22–32: Καὶ γὰρ ἐν τῷ διαλέγεσθαι φύσει ἡ φωνὴ καθ ’ ἑκάστην τῶν ξυλλαβῶν πρῶτόν τι καὶ δεύτερον τῶν γραμμάτων τίθησι καὶ τρίτον καὶ τέταρτον καὶ κατὰ τοὺς λοιποὺς ἀριθμοὺς ὡσαύτως , οὐ πᾶν μετὰ πᾶν, ἀλλ ’ ἔστι τοιαύτη τις φυσικὴ αὔξησις τῆς συνθέσεως . παραπλησίως δὲ καὶ ἐν τῷ μελῳδεῖν ἔοικεν ἡ φωνὴ τιθέναι κατὰ συνέχειαν τά τε διαστήματα καὶ τοὺς φθόγγους φυσικήν τινα σύνθεσιν διαφυλάττουσα , οὐ πᾶν μετὰ πᾶν διάστημα μελῳδοῦσα . Transl. Porter (2010). See also Pl. Soph. 253a–b, Aristox. Harm. 37.4–6 and Rhythm. 2.8, 4.27–30 make similar points. Cf. the discussion in Rocconi (2010) 180–181, who signals parallels between the passage quoted above and Dion. Hal. Comp. 11.8–10, quoted below. 51 Dion. Hal. Comp. 2.1. According to De Jonge (2008) 51, Dionysius combines the Stoic phrase ‘elements of language’ ( στοιχεῖα λόγου ) with a rhetorical approach to language as expression ( λέξις ). 52 Dion. Hal. Comp. 11.13–14: Μουσικὴ γάρ τις ἦν καὶ ἡ τῶν πολιτικῶν λόγων ἐπιστήμη τῷ πόσῳ διαλλάττουσα τῆς ἐν ᾠδαῖς καὶ ὀργάνοις , οὐχὶ τῷ ποιῷ. καὶ γὰρ ἐν ταύτῃ καὶ μέλος ἔχουσιν αἱ λέξεις καὶ ῥυθμὸν καὶ μεταβολὴν καὶ πρέπον , ὥστε καὶ ἐπὶ ταύτης ἡ ἀκοὴ τέρπεται μὲν τοῖς μέλεσιν , ἄγεται δὲ τοῖς ῥυθμοῖς, ἀσπάζεται δὲ τὰς μεταβολάς , ποθεῖ δ’ ἐπὶ πάντων τὸ οἰκεῖον , ἡ δὲ διαλλαγὴ κατὰ τὸ μᾶλλον καὶ ἧττον . Dion. Hal. Comp. 11.15–21 signals the differences between musical melody and the melody of language (cf. Aristox. Harm. 1.18.9–29); in addition, Dion. Hal. Comp. 11.25 and Dem. 50.7 argues that prose should be ‘rhythmical’

163

CHAPTER FOUR

Dionysius also refers to the science of music in order to illustrate the importance of properly combining one’s words: in the following anecdote, the critic aims to show that the natural law of arrangement is ingrained in all human beings, even in those with no technical knowledge whatsoever (i.e., rule 2 listed above). 53

Ἤδη δ’ ἔγωγε καὶ ἐν τοῖς πολυανθρωποτάτοις θεάτροις , ἃ συμπληροῖ παντοδαπὸς καὶ ἄμουσος ὄχλος , ἔδοξα καταμαθεῖν, ὡς φυσική τις ἁπάντων ἐστιν ἡμῶν οἰκειότης πρὸς ἐμμέλειάν τε καὶ εὐρυθμίαν , κιθαριστήν τε ἀγαθὸν σφόδρα εὐδοκιμοῦντα ἰδὼν θορυβηθέντα ὑπὸ τοῦ πλήθους , ὅτι μίαν χορδὴν ἀσύμφωνον ἔκρουσε καὶ διέφθειρεν τὸ μέλος , καὶ αὐλητὴν κατὰ τῆς ἄκρας ἕξεως χρώμενον τοῖς ὀργάνοις τὸ αὐτὸ τοῦτο παθόντα ὅτι , ἀσύμφωνον ἐμπνεύσας ἢ μὴ πιέσας τὸ στόμα , θρυλιγμὸν ἢ τὴν καλουμένην ἐκμέλειαν ηὔλησε . Καίτοι εἴ τις κελεύσειε τὸν ἰδιώτην τούτων τι ὧν ἐνεκάλει τοῖς τεχνίταις ὡς ἡμαρτημένων αὐτὸν ποιῆσαι , λαβόντα τὰ ὄργανα , οὐκ ἂν δύναιτο . τί δή ποτε; ὅτι τοῦτο μὲν ἐπιστήμης ἐστιν , ἧς οὐ πάντες μετειλήφαμεν , ἐκείνο δὲ πάθους , ὃ πᾶσιν ἀπέδωκεν ἡ φύσις . Τὸ δ’ αὐτὸ καὶ ἐπὶ τῶν ῥυθμῶν γινόμενον ἐθεασάμην , ἅμα πάντας ἀγανακτοῦντας καὶ δυσαρεστουμένους , ὅτε ἢ κροῦσιν ἢ κίνησιν ἢ μορφὴν ἐν ἀσυμμέτροις ποιήσαιτο χρόνοις καὶ τοὺς ῥυθμοὺς ἀφανίσειεν .

Before now I have thought I perceived, even in the most popular theatres, filled with a crowd of men of all kinds and of little culture ( παντοδαπὸς καὶ ἄμουσος ὄχλος ), how all of us feel naturally at home ( φυσική τις οἰκειότης ) with tuneful melody and good rhythm. I have seen an able and very renowned harpist booed by the public because he struck a single false note and so spoiled the melody. I have also seen a reed-pipe player who handled his instrument with supreme skill suffering the same fate because he blew thickly, or through not tightening his embouchure produced a discordant sound or what is called a ‘broken note’ as he played. And yet if anyone told the unskilled listener ( ἰδιώτης ) to take up the instrument himself and play any of the passages whose performance by professionals ( οἱ τεχνῖται ) he was criticizing, he would be unable to do so. Why ever is this? Because the latter is a matter of technical knowledge ( ἐπιστήμη ), while the former is a matter of sensation (πάθος ), which nature has conferred upon all men ( ὃ πᾶσιν ἀπέδωκεν ἡ φύσις ). I have observed the same

(εὔρυθμος ), that is, neither ‘in rhythm’ ( ἔνρυθμος ) nor wholly unrhythmical ( ἄρρυθμος ); cf. Arist. Rh. 3.8.1–3, and section 1.7 n. 151 above. See also section 2.3.3 on ancient approaches to rhythm in Demosthenes. 53 Dion. Hal. Comp. 11.8–10. Cf. Cic. De or. 3.195–197 (discussed below) and Orat. 173 (cited in n. 59 below).

164

OFFENDING THE EARS

thing occurring in the case of rhythms: everyone with one voice expresses annoyance and displeasure when a performer strikes an instrument, takes a step or makes a gesture out of time, and so destroys the rhythm.

Thus, if a musician makes even the tiniest mistake in arranging his melodies or rhythms, he ends up being booed and jeered at by his audience: the orator runs the same risk, mutatis mutandis, in arranging his words. 54 After all, Dionysius considered oratory a kind of music, containing rhythm and melody, with which every human being feels naturally at home; on a similar note, the author of On the Sublime describes word arrangement as ‘a certain melody of words which are part of man’s nature and reach not his ears only but his very soul’. 55 Dionysius and Longinus, then, believe that every human being has an inborn sensitivity to the arrangement of words. Therefore, experts rely on the same criteria as the average man on the street in the evaluation of smooth and rough collocations: Dionysius simply refers to the verdict of his ‘ears’ ( ἀκοαί ), his ‘sensation’ (πάθος ) or to his ‘irrational perception’ (ἄλογος αἴσθησις ). 56 The anecdote about the heckled musician also serves to show, however, that the uncultured masses lack the ‘technical knowledge’ ( ἐπιστήμη ) of the artist or the critic: hence, only the artist is capable of creating well-designed compositions, and only the trained scholar is able to substantiate why a given arrangement is smooth or rough (i.e. rule 3 listed above). Roman rhetoricians do not discuss the analogies between rhetoric and music as openly and as confidently as their Greek colleagues: as we have seen above, they take care not to be associated with unmanly or un-Roman affectations (section 4.2). 57 Still, Cicero underlines the

54 Cf. Cic. De or. 3.198, who thinks that orators will not suffer the same fate as the musician: he submits that orators are not heckled, when they make mistakes in the arrangement of their words, but instead, the audience silently takes offence at the poorly arranged words. The passage is discussed in the present section below. 55 Long. Subl. 39.3: Ἁρμονία τις λόγων ἀνθρώποις ἐμφύτων καὶ τῆς ψυχῆς αὐτῆς, οὐχὶ τῆς ἀκοῆς μόνης ἐφαπτομένων . Music and language have been viewed as related disciplines at least as early as the fifth century BC: Hippias of Elis worked ‘on the value of letters, syllables, rhythms and scales’ ( περί τε γραμμάτων δυνάμεως καὶ συλλαβῶν καὶ ῥυθμῶν καὶ ἁρμονιῶν, 86A11 Diels-Kranz). Cf. Porter (2010) 213–215. 56 For the judgment of the ear in Dionysius, see e.g. Comp. 23.20: ‘The irrational feeling of the ear testifies to it’ (τὸ ἄλογον ἐπιμαρτυρεῖ τῆς ἀκοῆς πάθος ). Dionysius refers to the effects of word arrangement on the ear on numerous occasions: see e.g. Dion. Hal. Comp. 10.2, 16.6, 16.8, 22.12–38; Dem. 40.3–10. Cf. also Demetr. Eloc. 48, 173–174; Philod. Poem. 5 col. 27.18–21 Mangoni, Rhet. 4 col. 2.2–10 p. 163 Sudhaus; Long. Subl. 39.4. Cf. n. 5 above on the ‘irrational criterion of perception’ in Dionysius; see n. 62 for Latin texts on the same topic. 57 Formarier (2011) argues that Cicero and Quintilian exaggerate the distinctions between delivering a speech and singing a song, so as to present their conservative ideal of the orator as an honest guardian of civic and moral

165

CHAPTER FOUR importance of word arrangement by referring to a story that is very similar to Dionysius’ theatrical tale: like the Greek critic, Cicero points out the universal human talent for recognizing rhythms and melodies, and he, too, addresses the critical faculties of both the uncultured masses and trained scholars. 58

Omnes enim tacito quodam sensu sine ulla arte aut ratione quae sint in artibus ac rationibus recta ac prava diiudicant; idque cum faciunt in picturis et in signis et in aliis operibus, ad quorum intellegentiam a natura minus habent instrumenti, tum multo ostendunt magis in verborum, numerorum vocumque iudicio; quod ea sunt in communibus infixa sensibus nec earum rerum quemquam funditus natura esse voluit expertem. Itaque non solum verbis arte positis moventur omnes, verum etiam numeris ac vocibus. Quotus enim quisque est qui teneat artem numerorum ac modorum? At in eis si paulum modo offensum est, ut aut contractione brevius fieret aut productione longius, theatra tota reclamant. Quid, hoc non idem fit in vocibus, ut a multitudine et populo non modo catervae atque concentus, sed etiam ipsi sibi singuli discrepantes eiciantur? Mirabile est, cum plurimum in faciendo intersit inter doctum et rudem, quam non multum differat in iudicando.

Everyone distinguishes what is good and bad in the systematic arts by means of a kind of inarticulate feeling ( tacito quodam sensu ), without the help of any art or system (sine ulla arte aut ratione ). People do so in the case of paintings, statues, and other works of art, for the understanding of which they are less well equipped by nature; and they display this capacity to a much greater degree when judging words, rhythms, and tones, because these are deeply rooted in our natural instincts ( in communibus infixa sensibus ), and nature has wanted no one to be entirely devoid of a feeling for such matters ( nec earum rerum quemquam funditus natura esse voluit expertem ). Consequently, not only artfully arranged words stir everyone’s feelings, but rhythms and the tones of the voice do so as well. After all, only a small minority understands the theory of rhythms and cadences; yet if the slightest mistake is made in these matters, and an element is contracted and becomes too short, or lengthened and

order. In distinguishing between music and oratory, Greek critics focused mainly on technical, not on moral issues: cf. n. 52 above. Yet, Aeschin. Ctes. 142 connects Demosthenes’ careful verbal arrangement to his allegedly deceitful conduct. 58 Cic. De or. 3.195–197, transl. May and Wisse (2001).

166

OFFENDING THE EARS

becomes too long, the entire theater cries out in protest. And surely the same thing happens with tones: choruses and even soloists, if they sing out of tune, are hooted off the stage by the ordinary crowd. Considering the great difference between the expert and the unschooled ( inter doctum et rudem ) in terms of performance (in faciendo ), it is remarkable how little they differ when it comes to making a judgment (in iudicando ).

Cicero reiterates these points in his Orator .59 The similarities between Cicero’s and Dionysius’ views on the evaluation of word arrangement are striking: their stories about slipping artists and booing audiences seem to belong to the standard repertoire of the Greek and Latin stylistic discourse in Rome. 60 Cicero also addresses a difference between theatrical and oratorical performances: ‘Just as the crowd discerns mistakes in verse, it also perceives if our speech limps in some way. But while they do not forgive a poet, they do make allowance for us; nonetheless, without being able to articulate it, all of them notice if what we have said is not well fitted and finished.’61 Cicero rightly concedes that the audience does not break into a vociferous protest, when an orator does not arrange his words neatly, but according to Cicero their silence does not make the judgment of the listeners less damning. Like Dionysius, Cicero always consults his ‘ear’ ( auris ), just as uneducated crowds rely on their ‘subconscious intuition’ ( tacitus sensus ), when assessing the quality of word arrangement. 62

59 Cic. Orat. 173: ‘Do they (i.e. those who reject the use of rhythm in oratory) never have the feeling that something is lacking, that a sentence is harsh, mutilated, lame or redundant? People do in the case of poetry, for the whole audience will hoot at one false quantity. Not that the multitude knows anything of feet, or has any understanding of rhythm; and when displeased they do not realize why or with what they are displeased. And yet nature herself has implanted in our ears the power of judging long and short sounds as well as high and low pitch in words’ ( nihilne eis inane videtur, nihil inconditum, nihil curtum, nihil claudicans, nihil redundans? in versu quidem theatra tota exclamant, si fuit una syllaba aut brevior aut longior; nec vero multitudo pedes novit nec ullos numeros tenet nec illud quod offendit aut cur aut in quo offendat intellegit; et tamen omnium longitudinum et brevitatum in sonis sicut acutarum graviumque vocum iudicium ipsa natura in auribus nostris collocavit ). 60 Donadi in Donadi and Marchiori (2013) 31–32 suggests that Dionysius ‘reproduces’ (‘ricalca’) Cicero’s passage. In my view, the similarities between the two texts can just as readily be ascribed to their participation in a common discourse, as I have also argued in in other cases: cf. section 1.2 n. 35 and 37 above. 61 Cic. De or. 3.198: Verum ut in versu vulgus, si est peccatum, videt, sic, si quid in nostra oratione claudicat, sentit; sed poetae non ignoscit, nobis concedit: taciti tamen omnes non esse illud, quod diximus, aptum perfectumque cernunt. Transl. May and Wisse (2001). Cf. Orat. 203. Yet, Cic. Orat. 168 notes he saw the audience burst into a cheer after hearing a happy cadence. 62 For the judgment of the ear in Cicero, see e.g. Orat. 159: ‘Speech should gratify the ear’ ( voluptati autem aurium morigerari debet oratio ). Cicero and Quintilian often use the ‘pleasure of the ear’ ( voluptas aurium ) and

167

CHAPTER FOUR

In sum, we have seen that both authors assume that there is a natural law for arranging words, which every human being instinctively understands, but which can only be fully explained by the likes of Cicero and Dionysius.

4.4 Disrupting the Natural Order: Sources for Roughness My next question is concerned with the sources for rough word arrangement: which types of collocation, according to Cicero and Dionysius, break the natural order and cause the ears to be vexed? In the table below I have listed the features that both authors (as well as their colleagues) discuss in their respective discussions of roughness. 63 The musical categories of ‘tone’ and ‘rhythm’ are used by Dionysius and Cicero themselves. Dionysius, for instance, lists them along with ‘variation’ ( μεταβολή ) and ‘appropriateness’ ( τὸ πρέπον ) as the four principal instruments of word arrangement. 64 Cicero, likewise, posits that ‘there are two things that please the ear, tone and rhythm’.65 I have added ‘sentence structure’ as a third category, as both men not only pay attention to combining words, but also to combining clauses and rounding off sentences. 66 the ‘judgment of the ear’ ( iudicium aurium ) as their criteria for judging arrangement: see e.g. Cic. De or. 3.169, 3.177, 3.183, Orat. 168, 172–173, 177–178, 198–199; Quint. Inst. orat. 9.4.114–118. The phrase ‘silent perception’ ( tacitus sensus ) is Cicero’s translation of ἄλογος αἴσθησις . Schenkeveld (1988) 304–305 suggests that Cicero has translated the λόγος in ἄλογος αἴσθησις as ‘word’ instead of ‘reason’: hence, ἄ-λογος is rendered as ‘without a word’, or ‘silent’. Schenkeveld argues that this translation is convenient for Cicero: while theatre audiences express their judgments by cheering or jeering, the orator’s public normally does not voice its opinion about the arrangement (cf. De or. 3.198, Orat. 203). 63 See esp. Cic. De or. 3.171–199, Orat. 149–236; Dion. Hal. Comp. 22, Dem. 38–39. 64 See Dion. Hal. Comp. 11.1: the categories constitute the four means to achieve the goals of word arrangement, that is, ‘beauty’ ( κάλλος ) and ‘pleasure’ ( ἡδονή ). On these two goals, see section 3.2.1 above and section 4.5 below. For μέλος as ‘tone’ instead of ‘melody’, see Costil (1949) 914: ‘Il ne s’agit pas de chant, de mélodie, mais de l’effet particulier produit sur l’oreille par la iunctura des mots ainsi que par la structure interne de ceux- ci et qui est due à la qualité sonore des phonèmes constitutifs.’ Cf. Dion. Hal. Comp. 11.13–14 (quoted in n. 52 above). 65 Cic. Orat. 163: Duae sunt igitur res quae permulceant auris, sonus et numerus. Like Dionysius, Cicero discusses the importance of variation (e.g. De or. 3.193, Orat. 215) and appropriateness (esp. De or. 3.210–212, Orat. 70–74). 66 Dion. Hal. Comp. 2.4–5 distinguishes between three levels of arrangement, viz. individual words, clauses (κῶλα ) and sentences ( περίοδοι ): for the arrangement of clauses and sentences, see Comp. 7–9, 22.4–6, 23.5–8. Cic. Orat. 149 and 201 lists the arrangement of clauses, which he subsumes under the heading of ‘balance’ (concinnitas ), as one of three topics of arrangement, together with tone (which he also refers to as compositio , cf. n. 26 above) and rhythm (numerus ). The latter category includes rounding off periods: see Orat. 168–171.

168

OFFENDING THE EARS

− Clashes of vowels ( concursus, hiatus ; σύγκρουσις ). Tone ( μέλος , sonus ) − Clashes of consonants. − Successions of long syllables. Rhythm ( ῥυθμός , numerus ) − Clauses and sentences that have no appropriate rhythmical conclusion ( clausula , βάσις ). − The length of the sentence is not adapted to the Sentence structure thought or to the speaker’s breath. − Lack of symmetry between the clauses.

Table 8: the sources for rough word arrangement according to Dionysius and Cicero

In interpreting the acoustic category of roughness, we should consider the following caveat: the table above outlines the commonly observed rules for the production of rough word arrangement, but it does not claim to say anything about actual Greek and Latin stylistic practice. It is generally agreed, for instance, that there is a considerable discrepancy between the rhythmical doctrines expounded in Cicero’s rhetorical treatises and the rhythmical habits exemplified in his own speeches. 67 Moreover, Jaana Vaahtera has shown that Dionysius’ views on euphony and word arrangement are only partially reflected in the passages that he cites in support of these views. 68 It speaks to the strength and durability of the common discourse on word arrangement that Greek and Roman authors alike subscribe to its core theses, even if their usefulness for the actual writing of prose is not always obvious. As we will see in the remainder of this section, Cicero and Dionysius (as well as their colleagues) link roughness to the occurrence of pauses in the flow of the sentence, caused by the effort that is required for the pronunciation of the words. 69

67 For an extended discussion of this discrepancy, see Bornecque (1907) 169–176. Several scholars, e.g., Laurand (1907) 152–171 and Schmid (1959), have tried to reconcile Cicero’s practices with his theories, few would deny the inconsistencies. The most famous problem concerns the so-called heroic clausula ( ― ᴗ ᴗ ― x): Cic. Orat. 217 seems to approve of it, but Zieli ński (1904) 163–170 shows that only 0,6% of the clausulae follow the pattern. Shipley (1911) claims that Cicero only considers clausulae heroic, if ictus and accent coincide. 68 Vaahtera (1997) demonstrates the arbitrariness of some of Dionysius’ criteria of roughness by applying them to the sample texts provided by him. In some cases, his criteria seem to hold up: the ‘rough’ passage from Thuc. (Comp. 22.34) contains more examples of hiatus than the ‘smooth’ passage from Isoc. ( Comp. 23.19), viz., 129 against 6. Yet, in other cases, theory and practice are at odds: the citation from Thucydides has fewer clashes of a final ν and a subsequent initial consonant than the quote from Isocrates, viz. 98 against 126: Cf. n. 73 below. 69 The idea that euphony and ease of pronunciation are linked can already be found in Aristox., e.g. Rhythm. 2.8 (σ is difficult to pronounce) and fr. 87 Wehrli ( ρ is easy to pronounce). See also Philod. Poem. 2 col. 22.4–23.1

169

CHAPTER FOUR

In their discussions of tone ( μέλος , sonus ), to begin with, Cicero and Dionysius zoom in on the juxtaposition of final letters and subsequent initial letters. 70 The Greek critic, for example, discusses the collocation ἐν χορόν from the opening line of a dithyramb by Pindar: ‘These letters cannot by their nature be combined and united, for it is not natural for ν to precede χ in the same syllable. Hence, when they form the boundaries between successive syllables, they do not produce a continuous sound, but there is bound to be a pause between the two letters, and this keeps their sounds distinct.’71 Similarly, in the opening line of Thucydides, Dionysius explains that the collocation Ἀθηναῖος ξυνέγραψε is rough, ‘since σ is never placed before ξ with a view to being pronounced with it in the same syllable: the sound of the σ must be arrested by a pause of silence before the ξ is heard’.72 According to Dionysius, it is the combination of a ‘semivowel’ (ἡμίφωνον ) and a ‘consonant’ ( ἄφωνον ) that creates the roughness in these cases: the same effect can be achieved through the ‘collision of vowels’ ( σύγκρουσις τῶν φωνηέντων ), or hiatus, which also creates a brief, but noticeable interval of time. 73 According to Dionysius, such pauses arise, because the consonants and vowels in question are pronounced in different parts of the mouth: ‘The

Janko on the view of Pausimachus of Miletus that ‘in general cacophony does not arise from any source other than difficulty in pronunciation’ ( καθόλου τὴν δυσηχίαν μὴ ἄλλοθεν ἢ ἐκ τῆς δυστομίας παραγείνεται ). 70 The fact that Dionysius refers to this topic with the musical term μέλος should not surprise us: as we have seen (n. 50 above), Aristox. Harm. 27.22–32 already pointed out the analogy between combining letters and arranging musical notes. Dion. Hal. Comp. 14.1 refers to Aristoxenus’ view (cf. fr. 88 Wehrli) that the letters of the alphabet can be divided into vowels ( φωναί ) and consonants ( ψόφοι ): cf. n. 73 below. 71 Dion. Hal. Comp. 22.15–16: Ἀσύμμικτα δὲ τῇ φύσει ταῦτα τὰ στοιχεῖα καὶ ἀκόλλητα· οὑ γὰρ πέφυκε κατὰ μίαν συλλαβὴν τοῦ χ προτετάχθαι τὸ ν, ὥστε οὐδὲ συλλαβῶν δύο μόρια γινόμενα συνάπτει τὸν ἦχον , ἀλλ ’ ἀνάγκη σιωπήν τινα γίγνεσθαι μέσην ἀμφοῖν τὴν διορίζουσαν ἑκατέρου τῶν γραμμάτων τὰς δυνάμεις . The text discussed by Dionysius is Pind. fr. 75 Schroeder. 72 Dion. Hal. Comp. 22.37: Οὐ γὰρ προτάττεται τὸ σ τοῦ ξ κατὰ συνεκφορὰν τὴν ἐν μιᾷ συλλαβῇ γινομένην· δεῖ δὴ τοῦ σ σιωπῇ καταληφθέντος τότε ἀκουστὸν γενέσθαι τὸ ξ. The text discussed by Dionysius is Thuc. 1.1.1, mentioned in the first paragraph of section 4.3 above. 73 Dion. Hal. Comp. 14 divides the letters of the alphabet in vowels ( ε, η, ο, ω, α, ι, υ), semivowels ( λ, μ, ν, ρ, σ, including the ‘double letters’ ζ, ξ, ψ) and consonants ( π, β, φ, τ, δ, θ, κ, γ, χ). Vaahtera (1997) 589 lists the combinations of consonants that Dionysius explicitly classifies as rough, namely ν followed by a consonant or semivowel, ς followed by ξ, and ρ followed by ρ. On the collision of vowels, see esp. Dion. Hal. Dem. 38.3, where the author refers to the practice by metrical and musical writers of filling the gap ‘by inserting semivowels’ ( ἑτέρων παρεμβολῇ γραμμάτων ἡμιφώνων ), possibly referring to the ν ἐφελκυστικόν . On the harshness of hiatus in Greek theory, cf. e.g. Demetr. Eloc. 68–74, Philod. Rhet. 4 col. 2.2–10 p. 163 Sudhaus, Hermog. Id. 1.7 Rabe. Janko (2000) 173–178 and Calcante (2005) 46–50 offer overviews of the ancient ‘system of euphony’, based on both Greek and Latin texts.

170

OFFENDING THE EARS process of the mouth’s altering from one shape to another, that is neither akin to it nor like it, entails a lapse of time, during which the smoothness and ease of the arrangement is interrupted.’ 74 Like Dionysius, Cicero refers to the noticeable gaps that appear, when final and initial letters cannot be articulated continuously: especially hiatus ( concursus vocalium, hiatus ), in his view, yields ‘gaping sounds’ ( hiulcae voces ), as it ‘tears apart the vowels and the words’ (voces distrahere, verba diducere ). 75 In addition, Cicero remarks that the Roman orators and poets of yore used to drop a letter in order to smoothen the collision of consonants, as in dignu’ locoque (instead of dignus locoque ) or posmeridianus (instead of postmeridianus ). 76 Cicero is not as thorough in his analysis of the causes of rough tones as Dionysius, but his approach is certainly not at odds with the critical framework that the Greek scholar uses. This goes all the more for Quintilian, who gives us the following example of colliding letters: ‘Consonants also, and especially the harsher ones, clash violently where words meet, for example a final s with a following initial x.’ Interestingly, the combination that Quintilian adduces does not occur in Latin, but it is frequent in Greek: in fact, it appears in the Thucydidean phrase Ἀθηναῖος ξυνέγραψε that Dionysius cites as a typical example of harsh

74 Dion. Hal. Comp. 22.25: Ἐν δὲ τῷ μεταλαμβάνειν τὸ στόμα σχηματισμὸν ἕτερον ἐξ ἑτέρου μήτε συγγενῆ μήτε παρόμοιον ἐμπεριλαμβάνεταί τις χρόνος , ἐν ᾧ διίσταται τὸ λεῖον καὶ εὐπετὲς τῆς ἁρμονίας . Cf. Quint. Inst. orat. 9.4.34: ‘When the vowels which come together are different, the roughness of the combination depends on whether the mouth is differently or similarly shaped in forming them’ ( atque cum aliae subiunguntur aliis, proinde asperiores erunt prout oris habitu simili aut diverso pronuntiabuntur ). 75 Cic. De or. 3.172, Orat. 150, 152. Cic. Orat. 77 compares the orator who entirely avoids hiatus to a mason, ‘almost cementing together his words’ ( verba verbis quasi coagmentare ): cf. section 4.2 above on the architectural imagery in Lucilius and Dionysius, and section 4.5 below on the (im)possibility of hiatus in the Latin language according to Cicero. See also Rhet. Her. 4.18: ‘We shall avoid the frequent collision of vowels, which makes the style harsh and gaping’ ( fugiemus crebras vocalium concursiones, quae vastam atque hiantem orationem reddunt ). Quint. Inst. orat. 9.4.33: when hiatus occurs, ‘the speech gapes, pauses, and, as it were, labors’ ( hiat et intersistit et quasi laborat oratio ). 76 In Orat. 153–164, Cicero professes his preference for anomaly over analogy: grammatical rules, in his view, must yield to the ‘custom which favors the ear’ ( consuetudo auribus indulgens ). Quoting examples from Ennius, Pacuvius, Accius, Lucilius, Terence and others, he argues that it is justified to adapt words so as to avoid an ugly sound. On the euphony and cacophony of consonants, see also Lucil. fr. 367–370 Krenkel (= 389–392 Warmington) on the cacophony of the letter r and the euphony of the letter s (cf. n. 21 above), and Quint. Inst. orat. 9.4.37–40 on combining x with s (cf. n. 77 below), s with s, and m with vowels. Cf. also Varro fr. 113 Götz-Schöll and August. Dial. 6, who call the syllables crux and trux ‘harsh’ ( asperae ), and the syllables lana and luna ‘smooth’ ( leves ).

171

CHAPTER FOUR arrangement. As the Greek and Roman conceptions of the aural effects of word arrangement are largely in agreement, Quintilian can without reservation include a Greek example in his discussion of Latin word arrangement. It is not inconceivable that Quintilian, who is familiar with Dionysius’ works, has On the Arrangement of Words in mind in this particular passage. 77 With respect to rhythm, again, the fluency and continuity of the pronunciation function as the main criterions for distinguishing between ear-pleasing and ear-jarring combinations. According to Dionysius, rough harmony ‘favors long words with long syllables’, while smooth harmony ‘uses not the longest rhythms, but those of average length or somewhat shorter’. The accumulation of long syllables, Dionysius submits, ‘gives rise to delay and interruption in the composition’.78 Cicero, again, is not as precise in his rhythmical analyses as the Greek critic, but he agrees that long rhythms slow down the pace of the sentence: he considers the spondee, for instance, ‘rather heavy and slow’ ( hebetior et tardior ), having a ‘steady movement’ ( stabilis quidam gradus ). 79 Cicero pays special attention to rhythms that, like a final chord, round off periods or colons: ‘Since the ear is always awaiting the end and takes pleasure in it, this should not be without rhythm.’ In other words, a sentence without a closing cadence ( clausula , βάσις ) is harsh: according to Cicero, the period by C.

77 Quint. Inst. orat. 9.4.37: Ceterum consonantes quoque, earumque praecipue quae sunt asperiores, in commisura verborum rixantur, ut s ultima cum x prima . Quintilian mentions Dionysius at 3.1.17, 9.3.89 and 9.4.88. To my knowledge, this striking parallel between Quintilian and Dionysius has not been noticed before. 78 Dion. Hal. Dem. 38.1: Ὀνόμασι χρῆσθαι φιλεῖ μεγάλοις καὶ μακροσυλλάβοις . Dion. Hal. Comp. 23.6: Χρῆται δὲ καὶ ῥυθμῶν οὐ τοῖς μηκίστοις , μέσοις δὲ καὶ βραχυτέροις . Ibid. 22.22: Ἀναβολήν τε ποιεῖ καὶ ἐγκοπὴν τῆς ἁρμονίας . Dion. Hal. Comp. 17 associates long syllables with grandeur (and hence with roughness) and short ones with pleasure (and hence with smoothness): the former category includes the spondee ( ― ―), the molossus (― ― ―), the bacchius ( ― ― ᴗ) and the hypobacchius ( ᴗ ― ―), while the latter category features the pyrrhic (ᴗ ᴗ) and the tribrach ( ᴗ ᴗ ᴗ). On the ancient terminology of the tribrach, see section 2.3.3 n. 90 above. The connection between length and grandeur is also explicit in Demetr. Eloc. 40 and 117: the author attributes Thucydides’ dignity to a large extent to the ‘long syllables in his rhythm’ ( ἡ τοῦ ῥυθμοῦ μακρότης ), while he thinks that a sentence which consists exclusively of long syllables is ‘frigid’ ( ψυχρός ). The dactyl, often simply referred to as the ‘heroic’ foot (ἡρῷος , herous ), was also widely considered a source of grandeur in prose: see e.g. Arist. Rh. 3.8.4, Dion. Hal. Comp. 17.11–12, Long. Subl. 39.4. 79 Cic. Orat. 216. Cicero, who generally favors smooth arrangements, thinks that ‘not even the spondee is entirely to be discarded’ ( ne spondeus quidem funditus est repudiandus ), especially in short clauses: ‘The heaviness and sluggishness compensate for the small number of feet’ ( paucitatem enim pedum gravitate sua et tarditate compensat ). Quint. Inst. orat. 9.4.83 states that ‘the more time units the feet each occupy, and the more stability they receive from long syllables, the weightier they make the style’ ( quo quique sunt temporibus pleniores longisque syllabis magis stabiles, his graviorem faciunt orationem ).

172

OFFENDING THE EARS

Gracchus that ends on improbos probet (― ᴗ ― ᴗ ―) would end in a smoother way, if the words are so rearranged that the sentence ends on probos improbare (ᴗ ― ― ᴗ ― ―). 80 Likewise, Dionysius argues that the two initial periods from Thucydides’ history ending on the words τῶν προγεγενημένων and καὶ διανοούμενον respectively ( ― ᴗ ᴗ ᴗ ― ᴗ ―) do not have an appropriate rhythmical finish: in the quoted examples, the rhythm ‘appears to have no beginning and no conclusion, as if it were part of the period but not its end’. 81 Setting aside the tricky issue as to why one clausula rounds off the sentence while another leaves it unfinished, we can conclude that Cicero and Dionysius both consider the period that has no rhythmically marked ending rough-sounding. On a related note, both men assign a disproportionate length to the rough period: as Dionysius puts it, ‘another peculiarity of this style is that the periods are independent and simple, neither coming to an end simultaneously with the sense, nor calculated to suit the breathing of the speaker’. 82 Such

80 Cic. Orat. 199: Cum aures extremum semper exspectent, in eoque acquiescant, id vacare numero non oportet. See n. 46 above for the full quotation from C. Gracchus. I scan the final syllable of a period or colon as long: cf. section 2.4.2 n. 120 above. Cicero’s own version of the sentence ends on a ditrochee ( ― ᴗ ― x, which he calls a dichoreus ). At Orat. 212–215, he relates that he once saw C. Carbo end two successive periods on a ditrochee: thereupon, an enormous roar arose from the spell-bound crowd. Although Cicero associates this clausula with Asia and warns his readers not to overuse it, he seems particularly fond of it: Winterbottom (2011) 217 suggests that ‘by listening to certain orators in Rome and especially to Greek rhetores in Asia (and Rome too), he perhaps came to feel an unconscious preference for a final ditrochaic word’. Cic. Orat. gives seven examples of good clausulae, four of which are one-word ditrochees, viz., improbare (Cicero’s modification of Gracchus), persolutas (by Carbo), comprobavit (by Carbo) and aestimasti (by Crassus, ibid. 224). Cf. ibid. 232, where Cicero praises three clausulae from his own speech Corn. , i.e., mercatoresque superarunt and potuisse superari, ending on a paean and a spondee ( ― ᴗ ᴗ ᴗ ― ―), and Aegyptoque vicerunt , ending on a cretic and a spondee (― ᴗ ― ― ―). Incidentally, Cic. Orat. 217 claims that a final spondee cannot be distinguished from a final trochee, as the length of the final syllable is always indifferent. 81 Dion. Hal. Comp. 22.42: Ἀκόρυφός τις φαίνεται καὶ ἀκατάστροφος , ὥσπερ μέρος οὖσα τῆς δευτέρας , ἀλλ ’ οὐχὶ τέλος . Cf. Dion. Hal. Dem. 43.11. The clausulae quoted are the only ones that Dionysius discusses. He claims that smooth harmony uses pleasant, well-defined clausulae: see Comp. 22.5, Dem. 40.9. Cf. however ibid. 39.5, where Dionysius says that in rough sentences pleasing clausulae may also arise accidentally: in such cases, ‘the spontaneous gift of fortune is not rejected’ (τὸ συμβὰν ἐκ τῆς αὐτομάτου τύχης οὐκ ἀπωθεῖται ). 82 Dion. Hal. Dem. 39.4 · Ἔτι τῆς ἁρμονίας ταύτης οἰκεῖόν ἐστι καὶ τὸ τὰς περιόδους αὐτούργους τινας εἶναι καὶ ἀφελεῖς καὶ μήτε συναρπαζούσας ἑαυταῖς τὸν νοῦν μήτε συμμεμετρημένας τῷ πνεύματι τοῦ λέγοντος . Cf. Dion. Hal. Comp. 22.5 with n. 36 above on the absense of meaningless patch-words in rough word arrangement. Arist. Rh. 3.9.5 uses the speaker’s breathing to measure the maximum length of the period. Demetr. Eloc. 42–47 associates exceedingly long periods with grandeur, ‘as the author hardly allows any pause to himself and the listener’ ( ἐκ τοῦ μόγις ἀναπαῦσαι αὐτόν τε καὶ τὸν ἀκούοντα ). Cf. Long. Subl. 38.3–4 on the sublimity of long

173

CHAPTER FOUR sentences either end prematurely or they seem to go on forever: Cicero argues that the words in the former case seem ‘broken’ ( infracta ) and ‘minced’ ( amputata ), while the words in the latter case may leave the speaker gasping for breath, which has a jarring effect on our sensitive ears. 83 The internal structure of the sentence can also contribute to the roughness of the sound: according to Dionysius, ‘the members are combined in an unusual and individual way, and not as most people would expect or wish’. Thus, the speaker who aims at harshness avoids spectacular figures of speech that bring balance to the sentence. 84 Cicero’s view does not deviate: he advises the orator, who aims to please the ear, to add ‘symmetry’ ( concinnitas ) to his sentences through the use of Gorgianic figures. The rough sentence, by contrast, is free from such polish. 85 In conclusion, this section has demonstrated that Cicero, Dionysius and their colleagues largely adopt a common approach to the sources for rough arrangement: while smooth compositions are fluent and easy to pronounce, rough collocations are limping and impossible to enunciate continuously. The analysis of harsh tones, stumbling rhythms and ill- hyperbaton, ‘which makes the audience terrified of a total collapse of the sentence, and forces them from sheer excitement to share the speaker’s anguish’ ( εἰς φόβον ἐμβαλὼν τὸν ἀκροατὴν ὡς ἐπὶ παντελεῖ τοῦ λόγου διαπτώσει , καὶ συναποκινδυνεύειν ὑπ’ ἀγωνίας τῷ λέγοντι συναναγκάσας ). 83 Cic. Orat. 170: ‘Why should they prefer to let the sentence limp or stop short rather than keep pace with the thought?’ ( quid est cur claudere et insistere orationem malint quam cum sententia pariter excurrere ). Cic. De or. 3.181 submits that, while the length of the period is conventionally derived from the duration of a single breath, ‘it turns out to be so attractive that, even if someone had been endowed with unlimited breath capacity, we would still not want him to deliver his words in an unbroken flow. It was found that what gratifies our ears is precisely what is not only just bearable to human lungs, but actually easy to deliver’ ( ita est suave, ut, si cui sit infinitus spiritus datus, tamen eum perpetuare verba nolimus; id enim auribus nostris gratum est, quod hominum lateribus non tolerabile solum, sed etiam facile esse potest ). Cf. Rhet. Her. 4.18: a long period ‘does violence both to the ear of the listener and to the breathing of the speaker’ ( et auditoris aures et oratoris spiritum laedit ). Cf. also Quint. Inst. Orat. 9.4.124–125 for the same point. 84 Dion. Hal. Dem. 39.6: Τὸ περιττῶς καὶ ἰδίως καὶ μὴ κατὰ τὴν ὑπόληψιν ἢ βούλησιν τῶν πολλῶν συζεύγνυσθαι τὰ μόρια . Cf. Comp. 22.45: ‘There is also a great imbalance between the clauses, great unevenness in the periods, many novel figures of speech, and frequent neglect of grammatical sequence’ ( πολλὴ δὲ καὶ ἡ τῶν κώλων ἀσυμμετρία πρὸς ἄλληλα καὶ ἡ τῶν περιόδων ἀνωμαλία καὶ ἡ τῶν σχημάτων καινότης καὶ τὸ τῆς ἀκολουθίας ὑπεροπτικόν ). Demetr. Eloc. 13 describes the members of the disjointed style as haphazardly compiled stones: see n. 24 above. The flexible use of grammatical categories (e.g., case, number, mood) is often associated with grandeur and roughness: Dion. Hal. Comp. 22.6, Demetr. Eloc. 65–66, Long. Subl. 23–24. 85 Cic. Orat. 164–167 describes the smoothening effect of balance: through the use of (Gorgianic) figures of speech, such as antithesis and isocolon, ‘the sentence becomes rhythmical by its very nature, even if no rhythm is intended’ ( suapte natura numerosa sunt, etiamsi nihil est factum de industria ).

174

OFFENDING THE EARS balanced sentences is built, as we have seen, on the postulate of a universal law of composition, for which all humans are thought to possess a congenital sensitivity. Thus, the critics and rhetoricians can claim that their theories merely supply proof for something that everyone instinctively knows to be true.

4.5 Echoing Greatness: the Virtues of Roughness in Greek and Latin Prose In the previous sections, I have laid out the shared theoretical framework on which the extant Greek and Latin discussions of word arrangement are built (section 4.3), and we have seen that these texts agree in principle on the sources for rough acoustics (section 4.4). The next issue is concerned with the stylistic applications of harsh composition, as they are advertised by the Greek and Roman critics and rhetoricians: why, and under what conditions, do they advise the author of artistic prose to produce disharmonious tones and limping rhythms that bluntly violate the universal laws of arrangement? Interestingly, Cicero, Dionysius and their colleagues disagree considerably on the virtues of rough word arrangement: once again, the common discourse on stylistic theory will prove to be malleable, prone to be bent according to the specific needs and purposes of its users. It is the goal of the present section to establish the motivations that underlie the various views on stylistic roughness in Late-Republican and Augustan Rome: we will see that the divergent opinions that our authors put forward are not merely built on their personal tastes and agendas, but that they also illustrate the complex relationship between Greek and Roman views on prose style. The surviving ancient texts on word arrangement associate three principal virtues with crude acoustics. These favorable effects, which I will discuss below under the headings of ‘simplicity and sincerity’ (section 4.5.1), ‘grandeur and sublimity’ (section 4.5.2) and ‘archaism and the patina of antiquity’ (section 4.5.3), are closely interrelated: in each case, roughness is interpreted as a means to reinforce the austerity and seriousness of the ideas that the words express. 86 Smoothness, conversely, is usually regarded as an emblem of luxuriousness and frivolity.

1. Roughness can invest a text with an aura of simple and straightforward sincerity, the mark of an author who is not distracted by trifling stylistic issues.

86 Cf. section 4.1 n. 15 above on the imitative quality of language. Usher (1985), De Jonge (2008) and others refer to ‘rough arrangement’ ( σύνθεσις αὐστηρά ) as ‘austere composition’, stressing the austerity that is typically associated with ear-jarring combinations, not only in Dionysius, but also in Cicero and other sources.

175

CHAPTER FOUR

2. Roughness can add the impression of grandeur, weight and solemnity to the words, whose slow, colliding sounds imitate, as it were, the sublime dignity of the thought.

3. Roughness can evoke a sort of venerable antiquity, a pleasant blemish that can be compared to the green moldy ‘patina’ ( πίνος ) on ancient bronze statues, or to the slowly formed ‘incrustation’ ( χνοῦς) on timeworn objects.

In discussing each of these three major stylistic virtues of roughness, I will present evidence for a mutual exchange of ideas between Greek and Roman experts in rhetorical theory and literary criticism. To be brief, the surviving Latin discussions of word arrangement tend to focus on the virtue of sincerity, whereas Greek authors are more disposed to emphasize sublimity. Still, Greek and Roman authors do not operate in two mutually exclusive bubbles: we will see that Roman rhetoricians often react explicitly to Greek views, while the Greek critic Dionysius sometimes seems to react, albeit implicitly, to the ideas of his Roman coevals. The idea that roughness can conjure up a kind of archaic glory, lastly, is promoted by Dionysius, but it is also an important bone of contention in the polemic between Cicero and his Atticist opponents: indeed, this notion, which involves echoing, in an almost literal sense, the classic masterpieces of yore, is integral to the shared Greek and Roman discourse on style.

4.5.1 Simplicity and Sincerity The first of the virtues to be discussed is the impression of uncontrived simplicity and guileless sincerity that may seem to emanate from rough composition. This feature is often admired in Latin sources: after all, we have seen that Roman orators, who seemed to be overly meticulous in arrangement, risked to be accused of unserious, un-Roman and unmanly behavior (section 4.2). Among their Greek colleagues, only Dionysius touches cursorily on the simplicity of harsh arrangement, possibly in response to the Roman focus on the topic. Many Romans regarded roughness as a sure sign that a text was free from deceptive razzle-dazzle. The self-proclaimed Attic orators in the circle of Calvus, for one, deliberately produced ‘broken and choppy sentences’, because they considered the use of rhythm too much like a ‘trick to catch the ear’ ( insidiae ad capiendas auris ): hence, ‘they think that the only one who attains the Attic norm is he who speaks in a rough and unpolished style.’ 87

87 Cic. Orat. 170: see section 4.2 n. 35 above. Ibid. 28: Putant enim qui horride incultuque dicat, (…), eum solum Attice dicere. On the stylistic views of the Atticists and their controversy with Cicero, see esp. section 1.4

176

OFFENDING THE EARS

Quintilian, likewise, asserts that the orator who is caught striving for smooth rhythms, loses both his credibility and all of his emotional impact: ‘If a judge thinks the man has time to spare for this, he will neither believe him nor be moved by him to grief or anger.’ 88 Similar ideas about the connection between rough arrangement and artlessness can be found in the younger Seneca and in Tacitus: the former declares that ‘verbal symmetry ( concinnitas ) is not a manly ornament’, while the latter has the orator L. Vipstanus Messalla compare rough composition to a ‘shaggy toga’ ( hirta toga ) to be preferred over the ‘gay-colored garb of a courtesan’ ( fucatae et meretriciae vestes ). 89 To summarize, conspicuous smoothness and the concomitant attention to word arrangement were often mistrusted by Roman audiences. Cicero, too, refers to the appearance of artless straightforwardness emanating from the jarring collocations of rough harmony: in his Orator , he advises his readers, for instance, not to ‘cement their words together’ (verba verbis coagmentare ), ‘for the hiatus and clash of vowels have something agreeable (molle quiddam ) about it and show a not unpleasant carelessness ( non ingrata neglegentia ) on the part of a man who is paying more attention to thought than to words’.90 Yet, Cicero, as we have seen, famously prefers euphony over dissonance: the hesitant praise of roughness quoted here is but a minor concession in an otherwise ardent apology of his own sonorous, rhythmical and neatly structured periods. Indeed, the Roman scholar devotes the bulk of his treatise to the promotion of smoothness, almost invariably writing off the topics of rough arrangement as products of ‘weakness’ (imbecillitas ) and as signs of an utter ‘incapacity of coherent speech’ (infantia ). 91 Thus, above (on issue of chronology), section 3.4 above (on Cicero’s defense against the Atticists on the basis of the three-style formula) and section 5.6.1 below (on Calvus’ conception of Atticism). 88 Quint. Inst. orat. 9.4.143: Nec potest ei credere aut propter eum dolere et irasci iudex cui putat hoc vacare. Cf. section 4.2 n. 37 above. 89 Sen. Ep. 115.3: Non est ornamentum virile concinnitas. Tac. Dial. 26.1–3: Messalla (c. 45–c. 80 AD) complains that the orators of his days ‘produce the rhythms of stage-dancing’ ( histrionales modos exprimunt ) and that ‘most of them boast that their speeches can be sung and danced to’ ( iactant cantari saltarique suos commentarios ). The orator enters the scene in Dial. 14 to present a eulogy for the ‘old orators’ ( antiqui oratores ) of Cicero’s era: Calboli (2003) discusses the personality of this prominent orator of the first century AD. Cf. section 5.6.1 below on the virtue of masculinity in the Roman rhetorical discourse. 90 Cic. Orat. 77: Habet enim ille tamquam hiatus et concursus vocalium molle quiddam et quod indicet non ingratam neglegentiam de re hominis quam de verbis laborantis. These words are cited approvingly by Quint. Inst. orat. 9.4.37. The phrase ‘cementing together’ ( coagmentare ) refers to elision (cf. n. 94 below); Cicero draws the imagery from the realm of architecture: cf. Lieberg (1956) 464–465 and section 4.2 above. 91 Cic. Orat. 23, 236. On the latter passage, see section 3.4 n. 80 above. Cicero’s fondness of smoothness is apparent throughout Orat. : see, e.g., his introduction of the three topics of arrangement in Orat. 149.

177

CHAPTER FOUR

Cicero stands out as a dissonant voice in an overwhelming chorus of Roman authors who unconditionally praise the sincerity of rough composition: even his avid admirer Quintilian, as we have seen, criticizes him for the predominant smoothness of his oratory. 92 Still, it is worthwhile to have a closer look at Cicero’s heterodox view of roughness, particularly with respect to hiatus, as it pertains to the interaction between Greek and Latin views on word arrangement. After his initial conciliatory assessment, Cicero later in his Orator adopts a more inflexible approach to the collision of vowels: ‘The Latin language, indeed, is so careful on this point that no one is so boorish as to be unwilling to run vowels togethers ( vocalis coniungere ).’ Cicero adds that ‘we are not allowed to make a pause between vowels (distrahere voces ), even if we should wish to do so’. 93 It is generally assumed that the rhetorician means that in Latin prose, as in poetry, subsequent vowels at word junctions (including final m and initial h) do not actually clash, resulting in a pause, but that they are actually reduced to a single syllable.94 Despite his earlier praise of hiatus, then, Cicero now insists that it is thoroughly un-Latin. The ambivalence in Orator stems from the author’s pragmatic approach to Greek stylistic doctrines: on the one hand, Cicero relies heavily on the works of Greek predecessors, who are accustomed to praise vowel clashes, but he cannot, on the other hand, completely subscribe to their views, as they are at odds with his passionate call for smooth composition and with his polemic against the champions of sober, strident

92 See section 4.2 n. 37 above. 93 Cic. Orat. 150: Quod quidem Latina lingua sic observat, nemo ut tam rusticus sit quin vocalis nolit coniungere. Ibid. 152: Nobis ne si cupiamus quidem distrahere voces conceditur. 94 Most studies of vowel junctions in Latin prose conclude that such junctions were subject to elision in a similar way as in poetry: see esp. Zieli ński (1904) 616–617, Sturtevan and Kent (1915) and Riggsby (1991). The latter offers a discussion of the principal passages from Cicero, Quintilian and later grammarians which provide explicit rules for vowel junction in Latin prose. In addition to hiatus, there are several alternatives for dealing with vowel junction in Latin, most notably ‘elision’ (the complete loss of the final syllable of a word), ‘prodelision’ (the complete loss of the initial syllable of a word) and synizesis (the merging together of two syllables without alteration of the letters, like the final pair of vowels of Πηληϊαδεω in Hom. Il. 1.1 or Oilei in Verg. Aen. 1.41). Riggsby uses the word ‘elision’ as an umbrella term for all these processes ‘in which vowel junction is resolved by some kind of reduction’. It is clear that vowel reduction (in whatever way) was generally preferred over hiatus, although the latter is often presented as a valid option: cf. Rhet. Her. 4.18, Cic. Orat. 77, Quint. Inst. orat. 9.4.36. On the intricacies of reconstructing the phonology of vowel junction in classical Latin, see esp. Allen (1978) 78–83. So as to argue for the banishment of hiatus from the Latin language, Cic. Orat. 152 adduces the ‘slightly uncouth speeches’ ( orationes horridulae ) of the elder Cato and the poetry of Ennius: the former is said to have shunned hiatus altogether, whereas the latter is thought to exhibit it only once.

178

OFFENDING THE EARS arrangement.95 In the end, therefore, Cicero advises his readers not to follow the example of the Greeks: by allowing pauses between vowels, Roman orators apply Greek rules for creating harshness, but they break the rules of sound Latinity. Although Cicero is not altogether unambiguous about the stylistic viability of hiatus, his final advice to the Roman orator is quite straightforward: rely on your own Roman ears in assessing the sound of prose and ‘let the Greeks figure it out for themselves’ (Graeci viderint ). 96 In other words, Cicero calls into question the compatibility of Greek scholarship on hiatus with Latin literature: while Greek critics often extol the virtues of vowel clashes, the Roman author subjects the prose of his mother tongue to a different law. The author of Rhetorica ad Herennium , incidentally, appears to share this view: as we have seen, he emphatically problematizes the applicability of Greek teachings on Latin prose, and he indeed advises against ‘frequent collisions of vowels, which make the style harsh and gaping’ (crebrae vocalium concursiones quae vastam atque hiantem orationem reddunt ). 97 Thus, among Roman authors, hiatus could be seen as a hallmark of candid sincerity, but it could also be presented as a vicious sort of Grecism. Among the contemporary Greek sources, next, Dionysius occasionally associated hiatus and the other sources of rough word arrangement with simplicity and straightforwardness: he submits, for example, that harsh composition ‘aims to emphasize its own unstudied and simple character’ ( τὸ ἀνεπιτήδευτον ἐμφαίνειν θέλει καὶ ἀφελές ). 98 By and large, however, the extant Greek sources associate roughness with

95 We should also take into account the different contexts of Orat. 77 (where hiatus is praised) and ibid. 150–153 (where it is utterly condemned). In the former passage, Cicero specifically discusses the orator of the plain style, ‘whom some deem to be the only true Attic orator’ (quem solum quidam vocant Atticum ): in other words, Cicero makes allowances for the views of his Atticist opponents. In the latter passage, however, Cicero is entirely occupied by the issue of word arrangement and the apology of his own smooth composition. 96 Cic. Orat. 151 offers a brief account of the various Greek approaches to hiatus: he submits that Isocrates and Theopompus carefully avoid it, that Thucydides and Plato (whose Menex. is mentioned) often insert it intentionally, and that Demosthenes ‘generally’ ( magna ex parte ) does not use it. For Cicero’s review of the use of hiatus in Latin literature, see n. 94 above. Cf. section 4.5.2 below on Quintilian’s treatment of hiatus, which interacts with Greek scholarship in a different way than Cicero’s discussion. 97 Rhet. Her. 4.18. The author cites the following phrase as a vicious instance of an obtrusive series of vowel junctions: Bacae aeneae amoenissime inpendebant (‘the copper-colored berries hung most invitingly’). Riggsby (1991) 329 suggests that the emphasis in the passage may lay on the word ‘frequent’ ( crebras ): hence, the anonymous author may not be opposed to hiatus per se, but rather to the repetition of the phenomenon. On the author’s approach to Greek criticism, see Rhet. Her. 4.1–10 with section 1.4 above. 98 Dion. Hal. Comp. 22.5. Cf. Dem. 39.5: rough word arrangement ‘prefers a certain unaffected simplicity of construction, with mostly short phrases, imitating the artlessness of nature itself’ ( ἀποιήτως δέ πως καὶ ἀφελῶς

179

CHAPTER FOUR grandeur rather than with simplicity: the harsh sounds, which are interpreted as honest and humble by Roman ears, are perceived as sublime by Greek ears (section 4.5.2). Still, Dionysius does address the simple frugality of roughness: in fact, he discusses some objections against smooth, euphonious arrangement that run along similar lines as the ones that we have encountered in Latin sources. In On the Arrangement of Words , he invokes a fictitious aficionado of roughness, who censures Demosthenes for the effort that he invested in the arrangement of his words: ‘Was Demosthenes such a helpless creature, then, that when he was writing speeches, he laid out meters and rhythms beside him as his materials, as clay- modelers lay out their molds, and tried to fit his clauses into them, adjusting the word order this way and that, keeping careful watch on his longs and his shorts and taking great trouble over the cases of his nouns and the moods of his verbs and everything else affecting the parts of speech? An orator of his standing would cut a poor figure if he were to involve himself with such trappings and fripperies.’99 Dionysius does not agree with such criticisms: he adduces the cases of Isocrates, who ‘spent ten years over the composition of his Panegyricus ’, and of Plato, who ‘even at the age of eighty, never let off combing and curling his dialogues and re-plaiting them in every way’, to show that great zeal in arranging one’s words can result in brilliant prose. 100 Like Cicero, then, Dionysius goes to great lengths to defend Demosthenes’ meticulous composition. Clearly, the Greek critic was familiar with the ideas of fellow scholars who

καὶ τὰ πλείω κομματικῶς κατεσκευάσθαι βούλεται , παράδειγμα ποιουμένη τὴν ἀκατάσκευον φύσιν ). For Greek approaches to hiatus (in Demetrius, Philodemus, Dionysius), see also section 4.5.2 below. 99 Dion. Hal. Comp. 25.30: Ὁ Δημοσθένης οὖν οὕτως ἄθλιος ἦν, ὥσθ ’ ὅτε γράφοι τοὺς λόγους , μέτρα καὶ ῥυθμοὺς ὥσπερ οἱ πλάσται παρατιθέμενος , ἐναρμόττειν ἐπειρᾶτο τούτοις τοῖς τύποις τὰ κῶλα , στρέφων ἄνω καὶ κάτω τὰ ὀνόματα καὶ παραφυλάττων τὰ μήκη καὶ τοὺς χρόνους καὶ τὰς πτώσεις τῶν ὀνομάτων καὶ τὰς ἐγκλίσεις τῶν ῥημάτων καὶ πάντα τὰ συμβεβηκότα τοῖς μορίοις τοῦ λόγου πολυπραγμονῶν; Ἠλίθιος μεντἂν εἴη εἰς τοσαύτην σκευωρίαν καὶ φλυαρίαν ὁ τηλικοῦτος ἀνὴρ ἑαυτὸν διδούς . Dion. Hal. Dem. 51.2–52.6 returns to the refutation of the critiques of Demosthenes’ elaborate word arrangement: like Comp. 25.29–44, this passage attributes the criticism to anonymous detractors. 100 Dion. Hal. Comp. 25.32: Τὸν πανηγυρικὸν λόγον (...) ἐν ἔτεσι δέκα συνετάξατο . Ibid.: Τοὺς ἑαυτοῦ διαλόγους κτενίζων καὶ βοστρυχίζων καὶ πάντα τρόπον ἀναπλέκων οὐ διέλειπεν ὀγοδήκοντα γεγονὼς εἴη. For Isocrates’ protracted composition of Paneg. , cf. Long. Subl. 4.2 (= Timaeus T 23 Jacoby). To illustrate Plato’s zealous devotion to the arrangement of his words, Dionysius refers to the anecdote that a writing tablet was found after his death, containing the opening words of Resp. (i.e., κατέβην χθὲς εἰς Πειραῖα μετὰ Γλαύκωνος τοῦ Ἀρίστωνος ) in various orders: cf. Demetr. Eloc. 21, Quint. Inst. orat. 8.6.64 and Diog. Laert. 3.37. Denniston (1952) 41 analyses the style of the line, noting that ‘unstudied as this opening appears, the art that goes to the making of it is yet susceptible to analysis’. For a discussion of Comp. 25.29-44, cf. section 4.2 n. 43–45 above.

180

OFFENDING THE EARS favored the sober simplicity of rough word arrangement over the careful scrutiny exhibited by Demosthenes and others: all in all, it is not unthinkable that Dionysius’ apology of Demosthenes is directed, at least partly, against the views of some of his Roman coevals. His consistent unwillingness to mention Roman rhetoricians by name might explain why the roughness-loving detractors of Demosthenes in his work remain anonymous. 101 As I argued before (section 1.2), it is very likely (though impossible to prove beyond doubt) that Dionysius, who actively interacted with Roman scholars in Rome, engaged himself with the topics that were at the center of their attention.

4.5.2 Grandeur and Sublimity As I noted in the previous section, Greek critics associate roughness less often with sincerity and artlessness than with grandeur and sublimity. This can easily be demonstrated, if we briefly return to the topic of hiatus: while Roman rhetoricians connect hiatus almost invariably to the simple style, many of their Greek colleagues instead maintain that vowel clashes have very little to do with simplicity. Demetrius, for one, claims that the author of the ‘plain style’ ( χαρακτὴρ ἰσχνός ) should shun vowel concurrences, especially if they involve clashes of long vowels or diphthongs: ‘If there is any hiatus, we should have it between short vowels, or between a short and a long, or at any rate, shorts in some shape or form.’ According to the critic, hiatus is more appropriate in the grand style (χαρακτὴρ μεγαλοπρεπής ), especially if it occurs between long vowels and diphthongs. 102 Philodemus, in

101 As Dionysius speaks in general terms, it is unlikely that he directs his defense of Demosthenes against any specific critics or rhetoricians. According to Leo (1889) 286, the anonymous critics of Demosthenes were ‘undoubtedly Asianists’ (‘ohne Zweifel Asianer’): this is not helpful in identifying the unnamed scholars, as there did not exist a unified school of Asianism to which ancient critics and rhetoricians tended to subscribe voluntarily (cf. section 2.2 above and section 5.2 below). Kirchner (2005) 175–176 connects Dionysius’ apology of Demosthenes to the foregoing passage of Comp. 25.5, where the critic compares the topic under discussion (viz., the blurry boundaries between prose and poetry) is like the ‘Mysteries’ ( Μυστήρια ), for which certain ‘initiation rites of style’ ( τελεταὶ τοῦ λόγου ) must be observed; the critics of Demosthenes’ careful arrangement, conversely, are depicted as ‘uninitiated people’ ( βέβηλοι ), who ‘reduce the most serious subjects to ridicule through their own callowness’ ( εἰς γέλωτα λαμβάνουσι τὰ σπουδαιότατα δι ’ ἀπειρίαν ). 102 Demetr. Eloc. 207: Καὶ ἤτοι βραχέα συγκρουστέον βραχέσιν (...) ἢ βραχέα μακροῖς (...) ἢ ἁμῶς γέ πως διὰ βραχέων . For the appropriate forms of hiatus in the grand style, see ibid . 72–73. For the appropriateness of hiatus in the ‘forceful style’ ( χαρακτὴρ δεινός ), see ibid. 299–301. Demetrius’ point that concurrences of long sound units are more sublime than concurrences of short sound units corresponds with the notion that the style becomes grander, if the interruptions and pauses are longer: cf. Dion. Hal. Comp. 22.1–3, Long. Subl. 40.3, Quint. Inst. orat. 9.4.33–34 (with n. 115 below) and section 4.4 above. See, for this point, also Porter (2016) 232–234, who

181

CHAPTER FOUR addition, considers vowel collision ‘quite frigid’ ( ὑπόψυχρος ), referring to one of the principal vices associated with the failure to achieve grandeur.103 In a word, Greek critics generally regard word combinations that contain long intervals as instrumental in achieving sublimity —the longer the pauses, the grander the effect. Dionysius’ view is not different: he, too, connects hiatus and other audible interruptions in the flow of the words to such virtues as ‘beauty’ ( κάλλος ), ‘grandeur’ (μεγαλοπρέπεια ), ‘weight’ ( βάρος ), ‘solemnity’ ( σεμνολογία ) and ‘dignity’ ( ἀξίωμα ). 104 The extant Greek discussions of rough word arrangement are built on a shared premise, which has left few traces in Roman rhetorical theory, but which has been fundamental to Greek approaches to language and style, at least since the publication of Plato’s Cratylus —the notion that in language and literature sound should imitate substance. 105 Hence, if an author deals with such themes as danger, war or suffering, the arrangement should be rough; if his text, however, is characterized by relaxation, peace and joy, the words should run smoothly. According to Dionysius, Homer is a master of acoustic mimesis, always seizing the appropriate occasion to apply dissonance and roughness: the critic points out that the poet uses ‘the letters that are most difficult to pronounce’ ( τὰ δυσεκφορώτατα) in his description of the terrifying Gorgon on Agamemnon’s shield; that he introduces ‘clashes of syllables and delays in the rhythm’ ( ἀνακοπαὶ συλλαβῶν καὶ ἀναβολαὶ χρόνων ) in the passage about Achilles’ struggle with Scamander; and that he dwells on ‘the most unpleasant and ill- sounding letters’ ( τὰ ἀηδέστατα τε καὶ κακ oφωνότατα γράμματα ) in the verses that describe how the Cyclops crushes the heads of Odysseus’ companions. 106 In his analysis of Homer’s

notes that ‘Dionysius is using the aesthetics of the gap that will become the hallmark of the Longinian sublime’. Hermog. Id. also associates hiatus particularly with the sublimes genres of ‘greatness’ ( μέγεθος ) and ‘beauty’ (κάλλος ), but he submits that the ‘pure style’ ( ἰδέα καθαρά ), which is characterized by simplicity and clarity, is ‘not at all fussy about the collision of vowels’ ( μηδὲν περὶ συγρκούσεως τῶν φωνηέντων μικρολογουμένη ). 103 Philod. Rhet. 4 col. 2.2–10 p. 163 Sudhaus: ‘Collision of vowels is rather frigid, but sometimes not inopportune’ ( σύμπτω [σ]ις δ[ὴ φ]ωνηέν [των ] ἐστὶ μὲν [ὑπόψ ]υ[χ]ρο [ς], ἡ δ’ οὐκ ἄκαιρος ). Transl. Hubbell (1920). For frigidity as a neighboring vice to the grand style, see section 2.4.1 n. 110 and section 3.4 n. 96 above. 104 For the principal virtues of Dionysius’ rough type of word arrangement, cf. section 3.2.1, esp. table 6 above. 105 Calcante (2005b) recognizes a system of euphony that was based on imitation and onomatopoeia in various ancient critical sources (esp. Demetr, Eloc. , Philod. Poem. , Dion. Hal. Comp. , [Aristid.] Rhet. , Cic. Orat. and Quint. Inst. orat. ), which he traces back to Pl. Cra. In his discussion of mimicking arrangement, incidentally, Dion. Hal. Comp. 16.4 mentions Plato’s dialogue as the first work ‘on etymology’ ( ὑπὲρ ἐτυμολογίας ). 106 Dion. Hal. Comp. 16.7–13 argues that Homer, ‘the poet with the most voices of all’ ( πολυφωνότατος ἁπάντων ποητῶν), is a master in imitating the content of his poetry through the arrangement of his words:

182

OFFENDING THE EARS famous description of Sisyphus’ toil, lastly, Dionysius offers the following penetrating review of the mimicking force of word arrangement. 107

Πρῶτον μὲν ἐν τοῖς δυσὶ στίχοις οἷς ἀνακυλίει τὴν πέτραν , ἔξω δυεῖν ῥημάτων τὰ λοιπὰ τῆς λέξεως μόρια πάντ ’ ἐστὶν ἤτοι δισύλλαβα ἢ μονοσύλλαβα· ἔπειτα τῷ ἡμίσει πλείους εἰσὶν αἱ μακραὶ συλλαβαὶ τῶν βραχειῶν ἐν ἑκατέρῳ τῶν στίχων· ἔπειτα διαβεβήκασιν αἱ τῶν ὀνομάτων ἁρμονίαι διαβάσεις εὐμεγέθεις καὶ διεστήκασι πάνυ αἰσθητῶς, ἢ τῶν φωνηέντων γραμμάτων συγκρουομένων ἢ τῶν ἡμιφώνων τε καὶ ἀφώνων συναπτομένων· ῥυθμοῖς τε δακτύλοις καὶ σπονδείοις τοῖς μηκίστοις καὶ πλείστην ἔχουσι διάβασιν ἅπαντα σύγκειται . Τί δή ποτ ’ οὖν τούτων ἕκαστον δύναται; Αἱ μὲν μονοσύλλαβοί τε καὶ δισύλλαβοι λέξεις πολλοὺς τοὺς μεταξὺ χρόνους ἀλλήλων ἀπολείπουσαι τὸ χρόνιον ἐμιμήσαντο τοῦ ἔργου· αἱ δὲ μακραὶ συλλαβαὶ στηριγμούς τινας ἔχουσαι καὶ ἐγκαθίσματα τὴν ἀντιτυπίαν καὶ τὸ ψῦγμα καὶ ἡ τῶν τραχυνόντων γραμμάτων παράθεσις τὰ διαλείμματα τῆς ἐνεργείας καὶ τὰς ἐποχὰς καὶ τὸ τοῦ μόχθου μέγεθος· οἱ ῥυθμοὶ δ’ ἐν μήκει θεωρούμενοι τὴν ἔκτασιν τῶν μελῶν καὶ τὸν διελκυσμὸν τοῦ κυλίοντος καὶ τὴν τοῦ πέτρου ἔρεισιν .

First, in the two lines in which Sisyphus rolls up the rock (Od. 11.595–596), except for two verbs all remaining words in the passage are either disyllables or monosyllables (i.e., ἤτοι , ὅ, μέν , χερσίν , τε, ποσίν , λᾶαν , ἄνω , ποτί , λόφον ).108 Next, in each of the two lines, the long syllables are half as numerous again as the short ones (i.e., nine longs against six shorts).109 Then, all the words are so spaced as to advance

Dionysius not only shows how the poet imitates frightening or august situations through rough arrangement, but he also discusses instances of smooth collocations which are appropriate for gentle and pleasing scenes. 107 Dion. Hal. Comp. 20.13–14. The verses that are discussed in this passage are Hom. Od. 11.593–596: ‘And Sisyphus I saw there under his great yoke of pain, with both his arms hard-straining an enormous rock to move. Buttressing that boulder with his legs and both his hands, he heaved toward the summit of the hill’ ( καὶ μὴν Σίσυφον εἰσεῖδον κρατέρ ’ ἄλγε ’ ἔχοντα , / λᾶαν βαστάζοντα πελώριον ἀμφοτέρῃσιν· / ἤτοι ὃ μὲν σκηριπτόμενος χερσίν τε ποσίν τε / λᾶαν ἄνω ὤθεσκε ποτὶ λόφον ). Dionysius cuts the last verse (11.596) off after the fourth foot. Cf. n. 113 below on Demetr, Eloc. 72, who discusses the same passage, and n. 114 below on Long. Subl. 40.4, who seems to echo Dionysius’ discussion of the passage. 108 Dionysius must refer to the verses 11.595–596, as the other two (11.593-594) contain several words with more than two syllables, viz., beside the verbs, Σίσυφον , πελώριον and ἀμφοτέρῃσιν . 109 In this case, Dionysius’ statement only holds good, if it applies to the first three verses, which scan as follows: ,― ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ ― ― ׀ ― ― (11.594) ,― ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ ― ― ׀ ᴗ ᴗ ― ׀ ― ― (11.593) respectively. The fourth line (11.596), however, has ― ― ׀ ᴗ ᴗ ― ׀ ― ― ׀ ᴗ ᴗ ― ׀ ― ― ׀ and (11.595) ― ᴗ ᴗ

183

CHAPTER FOUR

in ample measures, and the gaps between them are distinctly perceptible, either because of the coincidence of vowels (i.e., ἄλγε ’ ἔχοντα , ἤτοι ὁ, ἄνω ὤθεσκε ) or the juxtaposition of semivowels or voiceless letters (i.e., εἰσεῖδον κρατέρ ’, λᾶαν βαστάζοντα , σκηριπτόμενος χερσίν τε , ποσίν τε );110 and the dactylic and spondaic rhythms are the longest possible and take the longest stride.111 Now what is the effect of each of these details? The monosyllabic and disyllabic words, leaving many intervals between each other, portray the long duration of the action; while the long syllables, which have a holding, delaying quality, portray the resistance, the weight and the difficulty. The drawing-in of breath between the words and the juxtaposition of rough letters indicate pauses in his efforts, the delays and the hugeness of his labor; and the rhythms, when considered in respect of their length, portray the straining of the limbs, his dragging effort as he rolls his burden, and the pushing upward of the stone.

After several rough verses, the Sisyphus passage concludes with a smooth line, appropriately containing a maximum number of dactyls, imitating the speed of the boulder as it rolls down from the top of the hill: ‘Do not the words’, Dionysius reacts, ‘when thus combined, tumble downhill together with the impetus of the rock? Indeed, does not the speed of the narration outstrip the rush of the stone?’ 112 Dionysius’ views about the mimetic quality of word

.ᴗ ᴗ. Cf. Aujac and Lebel (1981) 143 n. 2 ― ׀ ᴗ ᴗ ― ׀ ― ― ׀ five long syllables against six short ones: ― ᴗ ᴗ Note that I scan the final syllable of each verse as long: cf. section 2.4.2 n. 120 above. 110 For Dionysius’ division of letters into vowels, semivowels and voiceless letters, see section 4.4 n. 73. The number of hiatuses and clashes between semivowels and consonants is larger, if we take the instances of internal concurrences into account, viz., ἄλγε ’, ἔχοντα , λᾶαν , βαστάζοντα , ἀμφοτέρῃσιν , σκηριπτόμενος and ὤθεσκε . 111 Dion. Hal. Comp. 15.3–10 argues that ‘there is more than one kind of length and shortness of syllables; some are actually longer than the long and some shorter than the short’ ( μήκους καὶ βραχύτητος συλλαβῶν οὐ μία φύσις , ἀλλὰ καὶ μακρότεραί τινές εἰσι τῶν μακρῶν καὶ βραχύτεραι τῶν βραχειῶν). Dionysius compares, for instance, the quantity of the first syllable in the following words, in increasing order of length: ὁδός , Ῥόδος , τρόπος , στρόφος . In the Homeric passage under discussion, Dionysius may be thinking of the combinations such as μὴν Σ-, -δον κρ -, μὲν σκ -. Cf. Aujac and Lebel (1981) 143 n. 3. 112 Dion. Hal. Comp. 20.16: Οὐχὶ συγκατακεκύλισται τῷ βάρει τῆς πέτρας ἡ τῶν ὀνομάτων σύνθεσις , μᾶλλον δὲ ἔφθακε τὴν τοῦ λίθου φορὰν τὸ τῆς ἀπαγγελίας τάχος; The verse under discussion is Hom. Od. 11.598: ‘And downwards it hurtled, the pitiless boulder, rolling to the plain’ ( αὖτις ἔπειτα πέδονδε κυλίνδετο λᾶας ἀναιδής ), Dionysius notes that the verse contains no .― ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ ᴗ ᴗ ― ׀ which scans: ― ᴗ ᴗ monosyllabic words, that the majority of the syllables is short (viz., ten short against seven long), and that there are no instances of hiatus or clashes between semivowels and consonants at word junctions. The critic also

184

OFFENDING THE EARS arrangement are shared by Demetrius and Longinus. The former, for instance, notices how ‘in many passages, grandeur is produced by a series of ugly sounds’ ( δυσφωνία ); after quoting a verse from the Iliad about Ajax doing battle with Hector, he remarks that ‘in other respects the ugly clash of sounds is perhaps unpleasant to the ear (δυσήκοος ), but by its very excess it brings out the greatness of the hero, since in the grand style smoothness and euphony find only an occasional place’. 113 The author of On the Sublime , in addition, argues that a sublime idea ‘gains additional grandeur, from the fact that the rhythm is not hurried along, or, as it were, running on rollers (ἐν ἀποκυλίσματι ), but the words prop one another up (στηριγμοὺς ἔχειν ) and are separated by intervals, so that they stand firm and they give the impression of stable grandeur’ (πρὸς ἕδραιον διαβεβηκότα μέγεθος ). According to the critic, nothing is more destructive of sublimity than quick successions of short syllables, which resemble an effeminate ‘dance-rhythm’ ( ὀρχηστικόν ). 114 Among Roman authors, we do not find much discussion of acoustic imitation: as we saw, they connect roughness and smoothness to the moral integrity of the author rather than to the content of his narrative. Quintilian, however, combines the two approaches: his discussion recognizes instances of the ‘irrational long syllable’ ( μακρὰ ἄλογος ), which he presumes to be shorter than a ‘perfect long’ ( μακρὰ τελεία ): cf. Dion. Hal. Comp. 17.12 with the discussion of Ruijgh (1987). 113 Demetr. Eloc. 48 (on Hom. Il. 16.358): Ποιεῖ δὲ καὶ δυσφωνία συνθέσεως ἐν πολλοῖς μέγεθος . Ibid.: Ἄλλως μὲν γὰρ ἵσως δυήκοος ἡ τῶν γραμμάτων σύμπληξις , ὑπερβολὴ δ’ ἐμφαίνουσα τὸ μέγεθος τοῦ ἥρωος· λειότης γὰρ καὶ τὸ εὐήκοον οὐ πάνυ ἐν μεγαλοπρεπείᾳ χώραν ἔχουσιν , εἰ μή που ἐν ὀλίγοις . Demetr. Eloc. 72 discusses Hom. Od. 11.595 (the Sisyphus passage), noting that the double hiatus of long vowels in λᾶαν ἄνω ὤθεσκε ‘reproduces the stone’s upward movement and the effort needed’ ( μεμίμηται τοῦ λίθου τὴν ἀναφορὰν καὶ βίαν ). Demetrius’ analyses of hiatus are not strictly limited to the topic of word arrangement, as he explicitly takes internal hiatuses (as in λᾶαν ) into account; cf. esp. Eloc. 69–70. 114 Long. Subl. 40.4: Ἁδρότερον δὲ γέγονε τῷ τὴν ἁρμονίαν μὴ κατασπεῦσθαι μηδ ’ οἷον ἐν ἀποκυλίσματι φέρεσθαι , ἀλλὰ στηριγμούς τε ἔχειν πρὸς ἄλληλα τὰ ὀνόματα καὶ ἐξερείσματα τῶν χρόνων πρὸς ἑδραῖον διαβεβηκότα μέγεθος . See ibid. 41.1 for the incompatibility of sublimity and short-syllabled rhythmical feet, such as pyrrhics ( ᴗ ᴗ), tribrachs ( ᴗ ᴗ ᴗ) trochees ( ― ᴗ) and ditrochees ( ― ᴗ ― ᴗ): cf. section 2.3.3 n. 90 above for Dionysius’ view on these swift rhythmical feet. There are striking similarities between Dion. Hal. Comp. 20.13–14 (on the Sisyphus passage) and Long. Subl. 40.4: on the latter text, Russell (1964) notes that ‘this is the same sort of criticism as practiced by Dionysius’. Porter (2016) 407–408 points out several striking verbal correspondences: Longinus’ passage, for instance, refers to a ‘rock’ ( πέτρα ) and a ‘rolling-machine’ (ἀποκύλισμα ), which resonate with Dionysius’ discussion of Sisyphus who ‘rolls up his rock’ ( ἀνακυλίει τὴν πέτραν ). Moreover, both texts refer to words as ‘being a mighty stride’ (Long. διαβεβηκότα , Dion. Hal. διαβεβήκασι ), serving as ‘proppings’ ( στηριγμοί ) for each other. Hence, Porter can claim with confidence that ‘depending on their relative dates, it is extremely likely that Longinus is showing himself to be a close reader and emulator of Dionysius’.

185

CHAPTER FOUR of rough arrangement not only draws on the discourse of simplicity and sincerity, but he also touches on themes that are typical for the treatises of Demetrius and Dionysius. Like Demetrius, for instance, the Flavian teacher ranks several types of vowel collision according to the degree of their offensiveness; like Dionysius, he pays attention to the disposition of the mouth as it produces these various types of hiatus.115 Quintilian advises the orator-in-training not to shun the phenomenon altogether: ‘Hiatus is sometimes actually appropriate and adds a certain grandeur ( faciunt ampliora ).’ The rhetorician next quotes a corrupt phrase ( pulchra oratione acta [oratio ] iacta te ), noting that ‘the inherently long and rich syllables also take up a certain amount of time in the interval between the vowels, as though there was a pause there’. 116 Although I have distinguished between Greek and Roman approaches to word arrangement, Quintilian’s discussion of hiatus demonstrates that both perspectives could be combined as parts of a common stylistic discourse.

4.5.3 Archaism and the Patina of Antiquity To conclude the present discussion of the stylistic applications of rough word arrangement, I will now turn to the last of the three major recurring virtues of harshness, that is, its archaizing effect: the crudeness of the ear-jarring collocations could conjure up the rudimentary literary style of a long-gone epoch. 117 Apparently, roughness could be associated with primitive

115 Quint. Inst. orat. 9.4.33–34. See Demetr. Eloc. 72–73, 207 for the different degrees of offensiveness of hiatus; cf. n. 102 above. See Dion. Hal. Comp. 14.7–14 on the disposition of the mouth in pronouncing vowels; cf. section 4.4 n. 74 above. Note that Quintilian, unlike Cicero, does not insist that hiatus is forbidden in the Latin language, although he does mention the possibility of mitigating the harshness of hiatus by coalescing the two vowels into a single syllable ( συναλιφή ). For the plausibility of Quintilian’s familiarity with Dion. Hal. Comp. , see section 4.4 n. 77 above. 116 Quint. Inst. orat. 9.4.36: Nonnumquam hiulca etiam decent faciuntque ampliora quaedam. Ibid.: Longae per se et velut opimae syllabae aliquid etiam medii temporis inter vocales quasi intersistatur adsumunt. The hiatus- filled Latin phrase that Quintilian quotes is of unknown origin: I adopt the reading of Russell (2002), who follows Halm’s emendation iacta te for the word iactatae , which the mss have preserved . In this interpretation, oratio must be deleted. Russell proposes the following translation: ‘Show pride in yourself, having delivered a beautiful speech.’ 117 For the relationship between archaism and classicism, see esp. Kim (2014), who shows that Dionysius’ references to ‘archaism’ ( ἀρχαϊσμός ) and ‘the archaic’ ( τὸ ἀρχαῖον ) do not refer to a pre-Classical period (in fact, as we saw in section 1.6 above, neither Dionysius nor Cicero pay much attention to the period preceding the Classical era), but rather to a specific set of qualities that he considered fundamentally classical: ‘Archaic qualities, rather than occupying their own temporal sphere, are often located by Dionysius within the classical itself; for Dionysius, archaizing, somewhat paradoxically, is an essential part of the best classical writing.’ Cf.

186

OFFENDING THE EARS antiquity, while smoothness could be connected to refined modernity. This raised a controversial issue among the critics and rhetoricians in Rome: should the authors of their day imitate the rough-hewn acoustics that they recognized in several of their favorite authors from Classical Greece, even though that same literary canon contains many fine examples of exquisite euphony and sophisticated smoothness? The extant opinions on this topic are divided: while some praised rough arrangements of words as emblematic of rugged, venerable old age, others dismissed it as musty and outdated. We will see that Dionysius, Longinus and the Roman authors who called themselves ‘Attic’ advocated the former view, whereas Cicero vigorously defended the opposite position. In describing the archaizing effect of rough harmony, our authors often draw comparisons with the visual arts, that is, with timeworn sculptures and primitive paintings. Dionysius’ favorite metaphor for harsh arrangement is ‘patina’ ( πίνος ), referring to the moldy blemish that appears on the surface of bronze statues as the result of a prolonged exposure to the elements of nature. 118 This greenish layer can not only be seen as a lamentable symptom of decay, but it can also be thought to contribute to the austere venerability of the object: Plutarch, for instance, relates that a young visitor ‘marveled at the patina of the bronze’ of the statues of the Spartan commander Lysander and his officers in Delphi, ‘for it bore no resemblance to rust or verdigris, but it was smooth and shining with a deep blue tinge’.119

Porter (2006c) 327: ‘The point to see is that archaism is a perspectival feature as much as a stylistic and periodizing one, and that from the later postclassical perspective all earlier writing into the fourth century seems broadly archaic.’ 118 As a metaphorical term in literary criticism, the word πίνος is diversely translated. Van Hook (1905) 44 has ‘tinge of antiquity’ and ‘classical style’; Roberts (1910) gives ‘mellowing deposit, tinge of antiquity, flavor of archaism’, Van Wyk Cronjé (1986) offers ‘musty antiquity’, ‘patina’ or ‘tinge’; Aujac (1992) has ‘rouille, teinte de véstusté, d’archaisme, patine’. I will use ‘patina’ for πίνος , as do Usher (1974), (1985) and Donadi and Marchiori (2013), because this word aptly reflects the sculptural metaphor. Donadi (2000b) 56–57 suggests that the patina image was prompted to Dionysius by the state of decay of the old Doric temples in his day; Fornaro (2001) demonstrates the influence of Dionysius’ use of the word patina on the aesthetic theories of Winckelmann; Porter (2006c) 327–328 and Kim (2014) 380–382 discuss the patina metaphor as part of their study of ancient classicism and archaism (cf. the previous note). 119 Plut. Mor. 395b: Ἐθαύμαζε δὲ τοῦ χαλκοῦ τὸ ἀνθηρὸν ὡς οὐ πίνῳ προσεοικὸς οὐδ’ ἰῷ , βαφῇ δὲ κυάνου στίλβοντος . Plutarch uses the word ἀνθηρόν to refer to the pleasing patina of the statue; πίνος denotes the dirty rust on its surface. The passage refers to the group of thirty-eight statues, depicting the Spartan victors at the battle of Aegospotami (405 BC): cf. Paus. 10.9.7–9 and Falaschi (2017). The young man in Plutarch’s story thinks that the patina befits the statues of the admirals ‘as they stood there with the true complexion of the sea and its deepest depths’ ( οἷον ἀτέχνως θαλάττιοι τῇ χρόᾳ καὶ βύθιοι ἑστῶτες ).

187

CHAPTER FOUR

Moving from sculpture to literature, Longinus attributes to sublime texts a ‘charming patina’ (εὐπίνεια ) and a ‘bloom as on the surface of beautiful bronzes’ ( γάνωσίς τις ὥσπερ ἀγάλμασι καλλίστοις ).120 Dionysius uses the image of patina without referring explicitly to its original sculptural context: he describes rough arrangement as ‘having a beauty that consists in its patina of antiquity’ ( τὸν ἀρχαϊσμὸν καὶ τὸν πίνον ἔχουσα κάλλος ), conferring upon the text an ‘incrustation in the form of an ancient patina’ ( χνοῦς ἀρχαιοπινής ).121 The comparison between rough-hewn sculpture and rough literature was widespread in the Greek and Roman critical discourse: the image recurs in the works of Demetrius, Cicero and Quintilian.122 Undoubtedly, Dionysius presents the archaic flavor as a commendable feature of rough composition. 123 According to him, the patina of antiquity can be found particularly in the lyric poetry of Pindar, in the tragedies of Aeschylus, and in the historiography of

120 Long. Subl. 30.1. Cf. ibid. 36.3, comparing sublime literature to ‘the faulty Colossus’ ( ὁ Κολοσσὸς ὁ ἡμαρτημένος ), which he pits against Polyclitus’ Doryphorus: according to the author, the latter is admired for its accuracy ( τὸ ἀκριβέστατον ), whereas the former, like works of nature, is commended for its greatness ( μέγεθος ). De Jonge (2013) argues that Longinus’ Colossus refers to the golden statue of Zeus in Olympia, mentioned in Pl. Phdr. 236b, claiming that the critic stresses its faultiness so as to underline the archaism in Plato’s style. 121 Dionysius uses ‘patina’ ( πίνος ), ‘patinated’ ( πεπινωμένος ) and ‘having the patina of antiquity’ ( ἀρχαιπινής ): see Dion. Hal. Dem. 5.3 (quoted in Pomp. 2.1), 38.6, 39.7, 44.2, 45.4, Comp. 22.6. The word ‘incrustation’ (χνόος or χνοῦς) can refer to the light porous film on any object, e.g., the layer of sea-salt on Odysseus’ body after he landed on the island of the Phaeacians (Hom. Od. 6.226), the fine down on a flower (Theophr. Hist. pl. 2.8.4), bloom on fruit (Theophr. Caus. pl. 6.10.7), or the first fuzzy hairs on a young man’s chin (Dion. Hal. Dem. 51.7, Comp. 25.35). Usener and Radermacher consider the word χνοῦς in Dion. Hal. Dem. 5.3 a gloss for πίνος , but it probably belongs to the original text: Dem. 38.6 uses the expression ‘incrustation in the form of an ancient patina’ ( χνοῦς ἀρχαιοπινής ), combining both metaphors for rough, archaic style. Dion. Hal. Isoc. 3.6–7 differentiates between the style of Lysias and Isocrates by comparing them to different styles of sculpture. 122 Demetr. Eloc. 14: ‘So the older style has something of the sharp, clean lines of early statues, where the skill was thought to lie in their succinctness and sparseness’ ( διὸ καὶ περιεξεσμένον ἔχει τι ἡ ἑρμηνεία ἡ πρὶν καὶ εὐσταλὲς, ὥσπερ καὶ τὰ ἀρχαῖα ἀγάλματα , ὧν τέχνη ἐδόκει ἡ συστολὴ καὶ ἰσχνότης ). Cic. Att. 12.6.4 refers the slightly old-fashioned expression ‘pray’ ( quaeso ) as ‘having a charming patina’ ( εὐπινές ). Cf. ibid. 14.7.2, where he describes a letter from his son as ‘patinated’ (litterae πεπινωμέναι ), remarking that ‘the patina of the letter shows that he has learnt something’ ( πίνος litterarum significat doctiorem ); see also ibid. 15.16a and 15.17.2 on the patina of young Marcus’ letters. Quint. Inst. orat. 2.5.23 argues that the older orators should only be imitated ‘after the layer of uncouthness incident to that age is removed’ ( deterso rudis saeculi squalore ); later on, ibid. 12.10.7, the author compares older literature to the ‘stiff’ ( duriora, rigida ) statues of Callon and Hegesias (fl. c. 500 BC). Cf. also Porter (2006c) 325–326 n. 62 on the Antonine expression ‘color of old age’ ( color vetustatis ) in Fronto Ep. 2.19, 4.9; Gell. NA 10.3.15, 12.2.10, 12.2.12 and 12.4.13. 123 Yet, in diction, Dionysius does not consider archaism a virtue: see section 3.2.2 and Kim (2014) 368–370.

188

OFFENDING THE EARS

Thucydides. Although Dionysius does not explicitly say so, the patinated incrustation is probably not an original feature of these early rough-sounding works: the noble stain takes a long period of time to emerge, and is therefore only visible to later audiences.124 Why do harsh, ear-jarring collocations remind Dionysius specifically of the authors just listed? We should note that they are by no means older than the other authors in Dionysius’ classical canon: the antiquated style of Pindar, for instance, postdates the smoothness of Sappho, and the old-fashioned style of Thucydides is preceded by the ‘well-mixed’ harmony of Herodotus. Still, Dionysius invariably assigns a primeval aura to the aforementioned rough authors, as their works exhibit virtues that are more appropriate for old age than for youth, such as ‘weight’ ( βάρος ), ‘dignity’ ( ἀξίωμα ) and ‘solemnity’ ( σεμνολογία ). The prototypical representatives of smooth harmony, by contrast, are associated with such youthful attributes as ‘freshness’ ( ὥρα ), ‘charm’ ( χάρις ) and ‘pleasure’ ( ἡδονή ).125 To be brief, by applying the rules of roughness, the authors in Rome could, in a conspicuous way, exhibit their affinity with the distant classical past. Not only Greek authors, but also Roman writers could turn to harsh composition to approximate the classical literature of yore. Cicero reports that his opponents, the self-styled Attic orators who advocated the unadorned frugality of Lysias, shunned the rhythmical finish of smooth composition, because ‘this was not done by the ancients’. 126 In his Orator , Cicero compares his oratorical foes to the lovers of primitive painting: ‘Suppose they prefer archaic painting which used only a few colors to the perfection of modern art; must we, go back to the ancients and reject the moderns? They pride themselves on the names of their ancient models. Antiquity does carry authority in the precedents it furnishes, as old age does in respect of

124 Dion. Hal. Dem. 39.7 singles out Aeschylus, Pindar and Thucydides as models of ‘solemnity’ (σεμνότης ) and the ‘patina of antiquity’ ( ἀρχαῖος πίνος ). To these three names may be added the other ‘rough’ authors in Dion. Hal. Comp. 22.7; cf. section 3.3 table 7 and n. 60 above. Porter (2006c) 327 proposes that Dionysius regards the old masters themselves as already archaizing: ‘The implication that not only any later classical writer employing the austere or archaic style, such as Antiphon, Plato, Antimachus, and Demosthenes, but also its first exponents, Empedocles, Pindar, Aeschylus, and Thucydides, might themselves by archaizing and not archaic, is left teasingly open by Dionysius (or else simply allowed).’ Cf. Kim (2014) 378: ‘While one suspects that the old- fashioned quality Dionysius ascribes to the austere mode would be perceptible only to a later classicizing audience, proof is difficult to come by, because Dionysius (…) is concerned more to describe the effect on the listener (…) than to establish the writer’s awareness of such effects.’ 125 See Dion. Hal. Comp. 11.2–3 and section 3.2.1 table 6 above. 126 Cic. Orat. 168: Non erat hoc apud antiquos. On the stylistic views of the self-proclaimed Atticists, see sections 1.4 and 3.4 above and section 5.6.1 below.

189

CHAPTER FOUR years; and this authority has great weight with me.’ 127 Dionysius draws on a similar analogy between literature and painting, when he compares the styles of Lysias and his younger fellow orator Isaeus: ‘There are some old paintings which are worked in simple colors without any subtle blending of tints but clear in their outline, and thereby possessing great charm; whereas the later paintings are less crisply drawn but contain greater detail and a subtle interplay of light and shade, and are effective because of the many nuances of color which they contain. Now Lysias resembles the older paintings by his simplicity and charm, and Isaeus their more elaborate and more skillfully wrought successors.’ 128 The archaism of roughness and the modernity of smoothness are prominent themes in the polemic between Cicero and the advocates of harsh composition. The former reports that there was a group of orators in Rome who called themselves ‘Thucydideans’ ( Thucydidii ), whom he describes as a ‘new and unheard-of group of ignoramuses’ ( novum quoddam imperitorum et inauditum genus ), orators who quaintly imitated the historian’s convoluted syntax: ‘No one succeeds in imitating his dignity of thought and diction, but when they have spoken a few choppy, disconnected phrases (mutila quaedam et hiantia ), which they could have formed well enough without a teacher, each one thinks himself a regular Thucydides.’ 129

127 Cic. Orat. 169: Quid, si antiquissima illa pictura paucorum colorum magis haec iam perfecta delectet, illa nobis sit credao repetenda, haec scilicet repudianda? Nominibus veterum gloriantur. Habet autem ut in aetatibus auctoritatem senectus sic in exemplis antiquitas, quae quidem apud me ipsum valet plurimum. According to Plin. NH 35.50, the oldest painters used only four colors, viz., black, white, yellow and red. Cf. Quint. Inst. orat. 12.10.3, who compares the development of painting to that of literary style. In a similar way as Cicero criticizes the sober Atticizing orators, Quintilian censures the admirers of the paintings of Polygnotus (fl. 475–450 BC) and his father Aglaophon: ‘Their simple color still has its admirers, enthusiastic enough to prefer these rude objects, the beginnings, as it were, of the future art, to the greatest of the later masters. I take this to be a pretentious claim to superior understanding’ ( quorum simplex color tam sui studiosus adhuc habet ut illa prope rudia ac velut futurae mox artis primordia maximis qui post eos extiterunt auctoribus praeferant, proprio quodam intellegendi, ut mea opinio est, ambitu ). 128 Dion. Hal. Is. 4.1–2: Εἰσὶ δή τινες ἀρχαῖαι γραφαί , χρώμασι μὲν εἰργασμέναι ἁπλῶς καὶ οὐδεμίαν ἐν τοῖς μίγμασιν ἔχουσαι ποικιλίαν , ἀκριβεῖς δὲ ταῖς γραμμαῖς καὶ πολὺ τὸ χαρίεν ἐν ταύταις ἔχουσαι . Αἱ δὲ μετ ’ ἐκείνας εὔγραμμοι μὴν ἦττον , ἐξειργασμέναι δὲ μᾶλλον , σκιᾷ τε καὶ φωτὶ ποικιλλόμεναι καὶ ἐν τῷ πλήθει τῶν μιγμάτων τὴν ἰσχὺν ἔχουσαι . Τούτων μὲν δὴ ταῖς ἀρχαιοτέραις ἔοικεν ὁ Λυσίας κατὰ τὴν ἁπλότητα καὶ τὴν χάριν , ταῖς δὲ ἐκπεπονημέναις τε καὶ τεχνικωτέραις ὁ Ἰσαῖος . 129 Cic. Orat. 30–32: Huius tamen nemo neque verborum neque sententiarum gravitatem imitatur, sed cum mutila quaedam et hiantia locuti sunt, quae vel sine magistro facere potuerunt, germanos se putant esse Thucydidas. Cicero’s principal objection against the imitation of Thucydides is that the latter’s work is an unsuitable model for oratory: cf. section 1.6 n. 134 above. Cicero does not mention the names of these supposed

190

OFFENDING THE EARS

According to Cicero, these followers of the Athenian historian end up with a bizarrely anachronistic and needlessly old-fashioned style: ‘Are men so perverse as to live on acorns after grain has been discovered? Are we, then, to suppose that the diet of men could be improved by the assistance of the Athenians, but that their oratory could not?’ 130 In this comparison, the acorns stand for the rugged style of Thucydides: Cicero complains that the Thucydidean orators in Rome prefer this meager diet over the rich nutrition provided by the Attic grains, that is, by the smooth rhythms and well-polished periods of the likes of Isocrates and Demosthenes. As Cicero continues to argue, one cannot fault the old masters for their primitive roughness, but later generations should know better: 131

Nec ego id quod deest antiquitati flagito potius quam laudo quod est; praesertim cum ea maiora iudicem quae sunt quam illa quae desunt. Plus est enim in verbis et in sententiis boni, quibus illi excellunt, quam in conclusione sententiarum, quam non habent. Post inventa conclusio est, qua credo usuros veteres illos fuisse, si iam nota atque usurpata res esset; qua inventa omnis usos magnos oratores videmus. (…) Legi enim audivique nonnullos, quorum propemodum absolute concluderetur oratio. Quod qui non possunt, non est eis satis non contemni, laudari etiam volunt. Ego autem illos ipsos laudo idque merito quorum se isti imitatores esse dicunt, etsi in eis aliquid desidero, hos vero minime qui nihil illorum nisi vitium secuntur, cum a bonis absint longissime.

I do not demand from antiquity what it has not; rather I praise what it has, particularly because I judge their excellence of greater concern than their deficiency. There is, in fact, more good in words and ideas, in which they excel, than in a rhythmical sentence

aficionados of Thucydides, but the orator-cum-historian C. Asinius Pollio (76 BC–4 AD) was probably among them; the historian Sallust (86–ca. 35 BC) is also often associated with the group, although Cicero refers specifically to orators. See esp. the discussions of Roman Thucydideanism in Leeman (1955) 183–208 and (1963) 179–187. According to Bonner (1939) 83, there is an uninterrupted tradition from Cicero’s novum genus to the supporters of Thucydides in Dionysius’ era, such as Q. Aelius Tubero, the addressee of Dion. Hal. Thuc. : cf. De Jonge (2017). Yet, there is no real proof to establish a direct link between the Thucydides imitators of the late Republic and those of the Augustus era. On Thucydides’ convoluted syntax, see De Jonge (2008) 214–216. 130 Cic. Orat. 31: Quae est autem in hominibus tanta perversitas, ut inventis frugibus glande vescantur? An victus hominum Atheniensium beneficio excoli potuit, oratio non potuit? The passage refers to the myth that mankind had to make do with acorn, until Triptolemus sowed grain in Attic soil: see e.g. Ov. Met. 5.643–661. 131 Cic. Orat. 169, 171.

191

CHAPTER FOUR

ending, which they lack. The rhythmical ending was a later invention, which I believe the ancients would have used if it had been known and employed in their day. 132 We see that after its invention, all great orators employed it. (…) For I have read and listened to not a few orators whose style was almost perfectly rhythmical in its cadence; but those unable to attain to this are not satisfied with not being criticized; they even wish to be praised for their failure. I, on the other hand, praise precisely those whom they profess to imitate, and I am quite right in doing so, although I find something lacking in them; but I have scant praise for these moderns who imitate only the weak points of the ancients while they are far from attaining to their real merits.

This passage encapsulates the dispute between the proponents of archaic roughness and the supporters of newfangled smoothness in Late-Republican and Early-Imperial Rome. We have seen that both Greek and Roman authors participate in this discussion, drawing on a shared set of conceptual parameters: Cicero and Dionysius, for instance, both build their expositions about the aural aspects of artistic prose on analogies with the visual arts. Hence, their opinions on the seemingly trifling issue of word arrangement touches on the core of their aesthetic taste, for they not only evaluate crude Thucydides and smooth Isocrates, but they also compare the stiffness of archaic kouroi to the well-proportioned body of the Doryphorus, and they, likewise, pit the rigid drawings of Polygnotus against the painstakingly detailed paintings of Zeuxis and Parrhasius. In a word, rough word arrangement resonates with a major category in ancient Greek and Roman aesthetic thought: it appeals to the uncomplicated beauty of ancient art that can be played off against the complex artistic charm of modernity.

4.6 Conclusion This chapter has focused on one of the hottest issues in the critical discourse on prose style in Late-Republican and Early-Imperial Rome—the theory of word arrangement, which is concerned with the aural evaluation of literature. This topic, as we have seen, not only

132 According to Cic. Orat. 175–176, Thrasymachus invented rhythmical prose, and Gorgias invented balanced sentence structure: thus, Cicero traces two crucial aspects of smooth composition back to the second half of the fifth century BC. Yet, he submits that Thrasymachus’ style is ‘too rhythmical’ ( nimis numerose ) and that Gorgias used his invention ‘rather immoderately’ ( intemperantius ) and ‘too boldly’ ( insolentius ), being ‘too fond’ ( avidior ) of his own style. It was not until Isocrates, according to Cicero, that smoothness was applied with success, as Isocrates used rhythm ‘with greater skill’ ( scientius ) than Thrasymachus, and he added balance to his sentences ‘with greater restraint’ ( moderatius ) than Gorgias. Cf. section 1.6 n. 136 above.

192

OFFENDING THE EARS attracted enthusiastic students (e.g., Cicero and Dionysius), but it also gave rise to tenacious scolders (e.g., Philodemus and the so-called Roman Atticists). Concentrating on the extant views about stylistic roughness, dissonance and cacophony, this chapter has contributed to our understanding of the classicism that underlies the critical works of the respective authors: we have seen that harsh acoustics were considered intrinsic to the aesthetic experience of the classics. Although some ancient scholars preferred euphonious composition, all surviving discussions recognize ear-jarring crudeness as inextricably connected to an important portion of the classical Greek literary canon. Hence, the aspiring heirs of the old masters often consciously arrange their words in such a way as to provide discomfort instead of pleasure to the ears of their audience. By doing so, they could achieve several aesthetic effects: as we have seen, rough word arrangement could be seen as a sign of authentic simplicity, it could be associated with sublime grandeur, and it could evoke the rudimentary beauty of venerable old age. With respect to the complex relationship between Greek and Latin stylistic theory, next, this chapter has taught us at least three lessons. First, despite the obvious phonological and acoustic differences between the Greek and Latin languages, we have seen that Greek and Roman authors analyze the sound of prose on the basis of a shared theoretical framework: specifically, I have argued that there exists a remarkable agreement between them concerning the natural principles that govern the arrangement of words, and concerning the appropriate method of analyzing the acoustic effects of artistic prose. Secondly, this chapter has also shown that Greek and Roman appreciations of rough and smooth collocations differ considerably: Greek sources generally hold that the sound of the composition should imitate the content of the narrative, whereas Latin texts are preoccupied with composition as a means to reflect the authors’ moral character. In the Roman rhetorical tradition, the careful arrangement of words could make an orator vulnerable to the charge of Greek, soft and effeminate behavior. The third point, to conclude, is perhaps the most important: we have seen that Roman authors were aware of and reacted to Greek views on word arrangement, and vice versa. In Latin discussions, for one, the topic is often explicitly labeled as a Greek activity, while Cicero emphatically deviates from Greek theory on hiatus. Dionysius, conversely, seems to defend his favorite orator Demosthenes against the Roman detractors of careful word arrangement. Dionysius, whose views modern classicists typically connect to the works of other Greek critics (section 1.2), seems to have involved himself in a stylistic discussion that particularly concerned Roman orators and rhetoricians: this is not surprising, if we take into

193

CHAPTER FOUR account that Dionysius dedicates his major treatise On the Arrangement of Words to the young Roman aristocrat Metilius Rufus and that he was acquainted with various Roman scholars (section 1.5). The next chapter will focus on a topic, which links the Greek critic to yet another Roman rhetorical debate—the nature of Attic style and the political and moral virtues which it was thought to represent. We will see that Atticism became a hot item in Rome through the fierce polemic between the orators Calvus and Cicero (late 50s and early 40s of the first century BC), after which Attic style became a standard topic for both Greek and Roman scholars to engage with.

194

Chapter 5 TO BE ATTIC OR NOT TO BE ATTIC : THE FLUIDITY OF ATTICISM IN ORATORY , POLITICS AND LIFE

5.1 Introduction In the previous chapters, I have studied several major topics in the Greek and Latin discourse on prose style: in each case, we have seen that some features are universal to the common discourse, while others are particular to its various participants on account of their specific preferences, purposes and programs. The present chapter will demonstrate the same two crucial points with respect to the topic of Atticism: again, I will argue that the views of the critics and rhetoricians in Rome are built on a shared conceptual framework, while each author at the same time interprets elements from this common repertoire so as to suit their own goals and motivations. Specifically, we will see that there is a remarkable interplay between, on the one hand, the stylistic views of three prominent scholars, viz., Calvus, Cicero, and Dionysius, and, on the other hand, the contemporary political situation in Rome. There was hardly a greater compliment for an orator or prose author in Late- Republican and Augustan Rome than to be called ‘Attic’ (Ἀττικός , Atticus ). Our record of the city’s obsession with Attic style, now standardly referred to as Atticism, goes back to the middle of the first century BC, when a group of Roman orators, presumably led by C. Licinius Macer Calvus, started to present themselves as Attic orators (sections 1.4 and 3.4). These men censured Cicero, the leading orator of the day, for being not Attic enough and even ‘Asian’ (Ἀσιανός , Asianus, Asiaticus ), which prompted the latter to write extensively about the nature of Attic and Asian oratory in his Brutus, De Optimo Genere Oratorum and Orator , insisting that his own style was in fact as Attic as Athens itself. 1 After Cicero, Atticism remained a crucial feature in the discourse about prose style in Rome: under Augustus, for instance, the intricacies of Attic style were discussed by the Greek critics Dionysius of Halicarnassus and Caecilius of Caleacte. 2 In both Latin and Greek stylistic theory, the epithet ‘Attic’ was used as

1 Cic. Orat. 23 praises Demosthenes, Cicero’s principal model: ‘I believe that not even Athens itself was more Attic than he’ ( quo ne Athenas quidem ipsas magis credo fuisse Atticas ). The translation is mine. 2 On Caecilius’ works, which survive only in a limited number of fragments, see section 1.5 above. The Suda lists several works that touch on the topic of Attic oratory and style, viz., On the Stylistic Character of the Ten Attic Orators (Περὶ τοῦ χαρακτῆρος τὼν δέκα ῥητόρων ), How the Attic and Asian Styles Differ (Τίνι διαφέρει ὁ Ἀττικὸς ζῆλος τοῦ Ἀσιανοῦ) and Against the Phrygians (Κατὰ Φρυγῶν), which probably was a lexicon of ‘Asian’ diction ‘in alphabetic order’ ( κατὰ στοιχεῖον ). Cf. esp. section 1.5 n. 101 above.

195

CHAPTER FIVE a shorthand for literary excellence: ‘speaking in an Attic way’ (Attice dicere ), to paraphrase one of Cicero’s favorite maxims, was thought to be the same thing as ‘speaking in the best way’ (optime dicere ). 3 This chapter explores the various conceptions of Attic style in Rome in the first century BC. We will see that the term ‘Attic’ was not a technical term that referred to a single, clearly defined, type of style. Rather, it was a flexible slogan whose meaning was continuously renegotiated. On the one hand, Greek and Latin sources attribute roughly the same defining features to the style of classical Athens: discussions of Attic style evoke such notions as masculinity, purity, sanity and moderation. Yet, on the other hand, the authors in Rome each highlight different aspects from this web of associations, not only to suit their respective stylistic programs, but also to demarcate their position in contemporary society. Style, after all, is the man himself: by presenting their stylistic views as Attic, our authors also associated their political ideas and their moral character with the venerable name of Athens, the undisputed symbol of freedom, democracy, vigor, self-restraint, wisdom, and kindred virtues.4 Thus, Atticism was an effective strategy to assert oneself not only as an author or as a critic, but also as a citizen and as a man: indeed, to be Attic or not to be Attic, that was the question. After an assessment of previous scholarship on Atticism (section 5.2), I will explore the most salient recurring features in the discourse about the culture, politics and literature of Classical Athens in Rome (section 5.3). Subsequently, I will compare the metaphors that Calvus, Cicero and Dionysius use for Attic style, illustrating the flexibility of Atticism (section 5.4). The latter point will be further demonstrated by studying the individual conceptions of the notion in the three aforementioned authors, focusing first on Cicero and

3 See Cic. Brut. 291, Opt. gen. 12 and Quint. Inst. orat. 12.10.26. 4 The idea that style is the man himself is an ancient commonplace. Diog. Laert. 1.58 credits Solon with saying ‘that speech is the mirror of action’ ( τὸν λόγον εἴδωλον εἶναι τῶν ἔργων ). Cic. Tusc. 5.47 (possibly referring to Pl. Resp. 400d) attributes another version of this idea to Socrates, who is said to have held ‘that, as is man himself, so is his speech’ ( qualis homo ipse esset, talem eius esse orationem ). Similarly, Sen. Ep. 114.1 has the following much-cited phrase, presented by the author as a Greek proverb: ‘Men’s speech is just like their life’ (talis hominibus fuit oratio, qualis vita ). See Mansfeld (1994) 186–191 for more ancient articulations of this idea. In modern usage, the maxim gained popularity through the French naturalist Georg Louis Leclerc, Comte de Buffon, who argued that ‘le style, c’est l’homme même’ ( Discours sur le style , August 25, 1753). Likewise, the Irish playwright Oscar Wilde wrote in a letter covering his newspaper submission ‘Fashions in Dress’ ( Daily Telegraph , 2 February 1891): ‘I don’t wish to sign my name, though I am afraid everybody will know who the writer is: one’s style is one’s signature always.’

196

TO BE ATTIC OR NOT TO BE ATTIC

Dionysius (section 5.5), whose views have already been discussed at length in the foregoing chapters of this dissertation, and next on the orator Calvus (section 5.6), whose opinions merit a closer look. The various conceptions of Attic style that we encounter in these instances are closely connected to contemporary political circumstances. We will see that Calvus’ Atticism, for one, is not only directed against Ciceronian eloquence, but also against the behavior of the triumvirs Caesar and Pompey (late 50s BC). Cicero, next, articulates his ideas about Attic style during the dictatorship of Caesar (48–44 BC): we will see that his conception of Atticism is as much a commentary on current oratorical trends as it is a grim warning against tyranny. Dionysius, lastly, comes to Rome after the fall of the republic: his Atticizing program ties in very well with the politics of the young emperor Augustus (27 BC–14 AD). Throughout the turbulent first century BC, Athens stood as a malleable icon not only of brilliant eloquence, but also of good government and impeccable ethics.

5.2 Atticism in Modern Scholarship As we saw, the term Atticism refers, in very broad terms, to the tendency in both Greek and Latin prose literature to admire and imitate the models of the Classical Athenian period, especially the Attic orators (section 1.6). 5 At this point, the reader may wonder why we need yet another discussion of this rather trite topic: it has, after all, been amply discussed in countless articles, books and lexical entries. In the last quarter of the nineteenth century, the issue of Atticism engaged such German philologists as Erwin Rohde, Georg Kaibel, Wilhelm Schmid, Eduard Norden, Ludwig Radermacher and Ulrich von Wilamowitz-Moellendorf.6 Yet, frustratingly, their debates were ridden with controversy and misunderstanding, which have trickled down into later discussions: the twentieth century saw a continuous quarrel about the origins, historical development and nature of Atticism, based largely on the problematic assumptions that were formulated by the philologists of imperial Germany. 7

5 For the relation between classicism and Atticism, see section 1.6 n. 120 above. 6 See, e.g., Rohde (1876) and (1886), and Kaibel (1885), who focus on the relationship between the notions of Atticism and Asianism on the one hand and the Second Sophistic (ca. 50–ca. 250 AD) on the other hand. Schmid (1887–1897) and Norden (1898) discuss the importance of Atticism in the wider context of the ancient history of Greek and Latin prose style. Radermacher (1899) zooms in on the origins and antecedents of Atticism, while Wilamowitz (1900) aims to write a synthesis of the foregoing literature. Goudriaan (1989) 595–677 offers an extensive survey of the debates of Wilhelmine Germany, providing illuminating expositions about the underlying motives and assumptions of the scholars concerned. 7 See Goudriaan (1989) esp. 601–606, who shows that the debates about Atticism in Germany were influenced to a large extent by contemporary discussions about educational and cultural politics (the so-called ‘Schulkrieg’):

197

CHAPTER FIVE

Jakob Wisse has appropriately compared the secondary literature to a dense, overgrown jungle: ‘Not only is a real consensus still difficult to reach, but also the various opinions are often hard to disentangle.’ 8 As modern scholarship often seems to obscure rather than illuminate the ancient phenomenon of Atticism, we have sufficient reason to take up the issue once again. A major problem that has frustrated modern scholarly discussions about Atticism, in my view, is this: many classicists tend to understand Atticism as a closed school of thought, a ‘movement’ with a long history, several subdivisions and many identifiable adherents. 9 In the disputes of the nineteenth century, this is the one idea that remained unchallenged. In Die antike Kunstprosa , for instance, Norden presents the entire history of style from the second century BC onwards as a battle between the rival schools of Atticism and Asianism, the archaizing ‘old style’ (‘der alte Stil’) versus the baroque ‘new style’ (‘der neue Stil’) respectively. 10 In his famous article ‘Asianismus und Atticismus’, however, Wilamowitz has according to Goudriaan, it is regrettable ‘that later generations have adopted the results of philological science in Wilhelmine Germany without sufficiently criticizing it’ (‘afkeurenswaardig is alleen dat latere generaties de resultaten van de filologische wetenschap van het wilhelmische Duitsland hebben gerecipieerd zonder er afdoende kritiek op uit te oefenen’). 8 Wisse (1995) 66–7. 9 See e.g. Leeman (1963) 142 (‘movement’), Dihle (1992) 1163 (‘Stilrichtung’), Hidber (1996) 42 (‘Bewegung’). We should be aware of the shortcomings of the terminology (‘movement’, ‘school’, etc.) that we inherited from the scholars of the late nineteenth and early twentieth centuries. Cf. Wisse (1995) 70–71: ‘When I call Atticism a ‘movement,’ that is only because I know no better term. It was in fact a movement, if we understand the term loosely. For what Cicero describes, and what Dionysius describes, is not a movement in the sense of a closed school of thought, with official members, and an official policy and programme.’ Wisse’s ‘loosely’ understood conception of the Atticist movement is still quite rigid, as he recognizes a ‘development’ of Atticism, with ‘stages’ and a ‘transmission’. In this dissertation, however, the designation ‘Atticist movement’ is reserved for the group of Roman ‘Attic’ orators, presumably led by Calvus, who were active between ca. 55 and ca. 45 BC: see the end of the present paragraph. I do not consider Dionysius, Caecilius and others to be exponents of any such movement. 10 Norden (1898) esp. 216–21. The binary scheme is, of course, an oversimplification. It was the basis, however, of the nineteenth-century debate about the antecedents of the Second Sophistic: according to Rohde (1876), for instance, the latter movement was a continuation of Asianism, while Kaibel (1885) argued that it was an offshoot of Atticism. Cf. Schmid (1887–1897): in the five volumes of Der Atticismus in seinen Hauptvertretern , he traces the history of Greek Atticism from Dionysius until Philostratus. In section 4.5.3 above, we have seen that the opposition between ‘old’ and ‘new’ style did in fact play an important role in the stylistic discussions of the first century BC. Yet, the categories of old and new style should not be conflated with the notions of Atticism and Asianism.

198

TO BE ATTIC OR NOT TO BE ATTIC already convincingly shown that Asianism rarely referred to a positive stylistic category: the label ‘Asian’ was simply a ‘slogan’ (‘Schlagwort’), a polemical term of abuse that few ancient authors willingly applied to themselves.11 Although the traditional conception of Asianism as a rigid stylistic denomination was thus debunked as early as the year 1900, it took all but another century before classicists gradually became aware that the ancient evidence does not support the thesis that there ever existed a rigid, centuries-spanning movement of Atticism. 12 Hence, the customary interpretation of Atticism dramatically overstates the importance of any alleged Atticist movement in Antiquity—I use the term ‘movement’ only for the close-knit and briefly flourishing group of self-styled Attic orators around Calvus (sections 1.4 and 5.6). Still, it has proved to be difficult to shake off the traditional, monolithic interpretation of Atticism. It will be instructive to see how scholars have dealt, for instance, with the disparate views of Calvus, Cicero and Dionysius, the three authors who are the focus of this chapter. The two Roman authors exchanged their conflicting stylistic views in the so-called ‘Atticist controversy’ of the late 50s and early 40s of the first century BC.13 As Cicero was the prime opponent of Calvus’ Atticist movement, he is usually considered an enemy of Atticism,

11 Wilamowitz (1900) esp. 4–8, 26; he does not challenge the notion of a large Atticist movement. For a rare instance of a positive use of the term ‘Asian’, see Sen. Controv. 10.5.21: ‘Craton, a very witty man and professed Asianist, who waged war with everything Attic’ ( Craton, venustissimus homo et professus Asianus qui bellum cum omnibus Atticis gerebat ). Cf. ibid for an example for Craton’s witty anti-Atticism. When the emperor Augustus gave Craton twenty-four sesterces, the equivalent of an Athenian talent, the orator said: ‘Either add something, or take something off, to stop it being Attic’ ( ἢ προσθὲς ἢ ἄφελ ’, ἵνα μὴ Ἀττικὸν ᾖ). On the basis of this passage, Luzzatto (1988) 238 concludes that ‘the label ‘Asian’ is a positive and generally acknowledged term denoting a specific style of declamation’ (‘l’etichetta “asiano” individua in positivo, e in modo condiviso dai diretti interessati, uno specifico stile declamatorio’). Huelsenbeck (2018) 164 n. 43, however, rightly objects that this cannot be inferred from the Craton passage alone: ‘It is conceivable that a speaker might embrace the term Asianism (…) as an act of reappropriation—i.e., the assertive co-optation of a derogatory term. This practice is suggested by modern example, e.g., reappropriation of the word “queer”.’ Note that Cicero’s treatment of Asianism is also not unequivocally negative: see section 1.6 above and section 5.6.1 n. 143 below. 12 See e.g. Gelzer (1979) 25–26: ‘Das sind alles sehr vage und allgemeine Bestimmungen, und sie bestätigen nur, dass “das Attische” nicht eine Bezeichnung für konkret definierte, feststehende Qualitäten oder gar für von Werken attischen Herkunft hergeleitete, sachliche oder stilistische Eigenschaften ist, sondern ein allgemeines Ideal, das als symbolische Qualitätsmarke den Werken und den Bestrebungen klassizistischer Gestaltungsweise und ihren Vorbildern zugelegt wird.’ Cf. n. 9 for Wisse (1995) 70–71. See also Whitmarsh (1996), quoted at the end of the present section, and Porter (2006b) 34–39 with section 1.6 on the flexibility of classicism. Douglas (1955), (1966) and (1973) already noted that the significance of Atticism ‘has been greatly exaggerated by modern scholars’; he focuses on Roman (not Greek) Atticism. 13 For the chronology of Calvus and his self-styled ‘Attic’ movement, see section 1.4 above.

199

CHAPTER FIVE tout court. 14 Calvus, conversely, who is the earliest Atticist on record, is often used as a touchstone for Roman Atticism: authors who, like Calvus, adopted a sober, unadorned style, are likely to be classified as Atticists, whereas authors who do not meet these standards, are usually excluded from the canon. Gualtiero Calboli, for example, lists T. Pomponius Atticus, C. Asinius Pollio, M. Iunius Brutus, Caesar and Sallust as Atticists, while he numbers Q. Hortensius Hortalus, Cicero and Marc Antony among the Asianists. 15 Furthermore, on the basis of Calvus’ purist program, scholars have associated ‘Roman Atticism’, whatever that phrase may refer to, with analogist theories of grammar and with Stoic views about simple style. 16 We should, however, not suspend our entire understanding of the Roman conception of Attic style upon Calvus: his movement (the only sure ‘Atticist movement’ on record) was, by all appearances, short-lived, with little tangible influence in the following decades. 17 In the

14 Cicero was considered ‘not Attic enough’ ( parum Atticus ) and ‘Asian’ ( Asianus ) by Calvus and his movement: see section 5.6.1 below on Quint. Inst. orat. 12.10.12 and Tac. Dial. 18.4. This view is often repeated in modern scholarship, explicitly in e.g. Norden (1898) 212–33, Leeman (1962) 91–111 and Calboli (1997a) 79, but also implicitly, as Cicero’s stylistic views are simply not classified under the heading of Atticism. 15 Calboli (1997a) 79–80 and (1997b) 262–263. Inscribing any author into Calvus’ Atticist movement is a tricky business, as our sources record virtually no direct links between Calvus and any other Roman authors: arguments are based largely on stylistic resemblances. Yet, the styles of the presumed Atticists listed above are quite disparate, differing in important respects from Calvus’ program. Leeman (1963) 136–167 proposes that Roman Atticism evolved in three phases, the first imitating Lysias (Calvus, Calidius, Brutus), the second following Thucydides (Asinius Pollio) and the third archaizing (Sallust, Cimber, L. Arruntius). The Atticist affiliations of Calidius and Caesar have been convincingly disproved: see Douglas (1955) and Garcea (2012) 49–77 respectively. Brutus was probably sympathetic to Calvus’ views, but that he was an Atticist in Calvus’ terms is by no means certain: see esp. Douglas (1966) xiii-xiv and Bowersock (1979) 63–64. As for the presumed Asianists listed above, they were all associated with Asian rhetoric by other ancient authors: for Hortensius, see Cic. Brut. 325; for Cicero, see e.g. Quint. Inst. orat. 12.10.12; and for Marc Antony, see Suet. Aug. 86. 16 For the connections between Atticism and analogist grammar, see esp. Dihle (1957) and (1992) 1166. The alleged connections are used as arguments to link Caesar, who composed a treatise On Analogy , to Calvus’ movement: see Hendrickson (1906) 97–98 and Garcea (2012) 49–77. For the link between Atticism and Stoicism, see Dihle (1957) 185 and Moretti (1990) 94–95. 17 See section 1.4 above; on the notion ‘Atticist movement’, see n. 9 above. Already in 45 BC, Cic. Tusc. 2.3 declares that the self-styled Atticists ‘have fallen silent’ ( conticuerunt ). I do not follow Douglas (1955) 241 and (1966) xiii, who argues that ‘the term “Atticist” should be reserved for the coterie centering on Calvus’, which means that Atticism only had a ‘temporary importance’ in Rome. While it is true that Calvus’ conception of Atticism is especially relevant in the stylistic debates of the late 50s and early 40s of the first century BC, the admiration of Attic style remained central in both Latin and Greek stylistic theory, long after Calvus’ death.

200

TO BE ATTIC OR NOT TO BE ATTIC brief span of their activity, Calvus and his followers did not create a definitive program of Atticism, but they did manage to provoke the great Cicero, instigating a lively and long- lasting debate about Attic style in Rome. As for Cicero, the fact that Calvus and others considered him thoroughly un-Attic should not exclude him from the present discussion of Atticism. After all, Cicero himself wrote about Attic style quite extensively and enthusiastically: if Atticism is not a fixed stylistic program but simply a tendency to admire and imitate classical Athenian literature, Cicero is an Atticist just as much as Calvus. Sure enough, Cicero offers a benign discussion of Asian style, and his conception of Attic style is a far cry from Calvus’ ideas—Cicero does not favor the simple style and his preferred model is Demosthenes rather than Lysias. 18 Yet, such discrepancies only underscore the flexibility of Atticism: the meaning of the term ‘Attic’ could be manipulated to accommodate diverse, even opposing views about style. Another major bone of contention in modern debates about Atticism is the relationship between the Greek and Roman conceptions of Attic style. Scholars generally suppose that there was a difference of scope dividing Roman and Greek authors: Calvus’ program, for instance, focuses on the simple style, purportedly accepting only Lysias and Hyperides as the true models of Atticism, whereas Dionysius, Caecilius, and later Greek authors adopted a much broader approach, taking virtually all classical Attic authors in consideration. 19 Therefore, some scholars have concluded that Roman and Greek Atticism are in fact two separate self-contained movements that have very little to do with each other. 20 Others, conversely, have tried to establish links between Roman and Greek views: taking into account that the earliest extant sources are actually in Latin, scholars have either argued that Roman Atticists somehow influenced Greek stylistic theory, or that the two groups build on an older,

18 For Cicero’s not unfavorable discussion of Asian style, see section 1.6 above and section 5.6.1 n. 143 below. Cic. Orat. 155 declares the author’s allegiance to anomalist grammar. Cicero’s preference for Demosthenes and his views on the simple style have been discussed at length above, esp. sections 2.3.2 and 3.4. 19 See for this analysis, e.g., Bowersock (1979) 57–9, Dihle (1992) 1169–1170, and Hidber (1996) 38–9. Yet, cf. n. 22 below on the possibility that Calvus also admired Demosthenes as a model of passionate oratory. 20 Douglas (1973) 125 n. 89 and Hidber (1996) 30–44, for instance, argue that Roman Atticism forms a coherent, self-sufficient unit that just happens to predate the extant Greek articulations of Atticism. According to Dihle (1977) the Greek and Roman views of Atticism ‘are only related to each other indirectly and through a different relationship with grammatical Atticism’ (‘hatten nur indirekt und zwar durch eine jeweils andere Beziehung zum grammatische Atticizmus miteinander zu tun’): according to him, Roman authors specifically turned to Greek grammar, aiming to apply its rules to the analysis of the Latin language, whereas Greek authors did not need to imitate the grammatical intricacies of Attic prose in detail.

201

CHAPTER FIVE now lost, Greek tradition. This issue is closely connected to the debate about the beginnings of Atticism: was the movement founded in the Greek world by Greeks, in Rome by Greeks, or did it actually start with the Roman orator Calvus and his entourage? 21 Such problems concerning the origin, evolution and transmission of Atticism are as yet unresolved. Any communis opinio on these issues is thwarted by the fact that concepts such as ‘Roman Atticism’ and ‘Greek Atticism’ are vague and misleading. As we have seen, the extant Roman conceptions of Attic style are exceedingly varied: Cicero, like many later Greek critics, already accepted a much wider range of Attic models than his rival Calvus, while other Romans reportedly imitated only Thucydides or Xenophon. 22 Likewise, we should not lump all Greek Atticists together: the conceptions of Atticism that we encounter in the works of Dionysius, Caecilius and the authors of the Second Sophistic are far from uniform, sometimes exhibiting striking parallels with Roman views. 23 It seems more plausible, therefore, that

21 Cf. section 1.2 above on the reluctance on the part of modern scholars to consider the possibility that Roman views could influence Greek theory. Norden (1898) 258–263, for one, posits that the extant Greek and Roman sources represent two independent offshoots of an originally Greek tradition, going back to the second century BC, while Wilamowitz (1900) 31–51 suggests that Atticism was founded by Greeks in Rome ca. 60 BC. More recently, Gelzer (1979) 15 connected Roman Atticism to Greek grammatical theory, and O’Sullivan (1997) 42 thought that ‘Caecilius was, if not the father of Atticism, at least present at its birth’; yet, on the probability that Caecilius did not flourish before the Augustan period, see again section 1.2 above. The possibility that Atticism is an originally Roman phenomenon was first considered by Radermacher (1899a) 360: ‘Es scheint mir hierbei durchaus unbewiesen, dass es nothwendig ein Grieche gewesen sein muss, der in Rom die Fahne des Streites erhob, so allgemein verbreitet auch heutzutage diese Annahme sein mag.’ Cf. also Kennedy (1972) 241–242, 351–353, Bowersock (1979) 67 and Innes (1989) 245–246. According to Heck (1917) 56, Dionysius’ Atticism is a ‘compromise between the Ciceronian and ultra-Atticist movement’ (‘Kompromiss der ciceronischen und der ultraattizistischen Richtung’). Wisse (1995) 74 offers the most confident claim about the Roman influence on Greek Atticism: ‘I think the movement was originally Roman, and was passed on to Greeks working in Rome.’ 22 For the presence of imitators of Thucydides and Xenophon in Rome, see Cic. Orat. 30–32. Additionally, Cic. Brut. 289 seems to react to Atticists who imitated Demosthenes. There is even some evidence that Calvus is among the admirers of Demosthenes: Sen. Controv. 7.4.8, for one, notes that ‘his forensic style is vigorous on the model of Demosthenes, with nothing sedate or gentle about it—everything excited and stormy’ ( conpositio quoque eius in actionibus ad exemplum Demosthenis viget: nihil in illa placidum, nihil lene est, omnia excitata et fluctantia ). Plin. Ep. 1.2.2 also mentions Calvus and Demosthenes together. Lebek (1970) 83–97 adduces the example of Calvus’ imitation of Demosthenes as evidence that his movement of Roman Atticists was not as homogeneous as is often assumed: similar observations have been made by Stroh (1982) 27–28, Narducci (1997) 130, and Dugan (2005) 311 n. 204. 23 Caecilius, for instance, seems to have taken a more grammatical approach than Dionysius: the list of his works in Suda includes two works that are listed as alphabetically arranged lexica, viz., Against the Phrygians (Κατὰ

202

TO BE ATTIC OR NOT TO BE ATTIC

Romans and Greeks participated in a shared discourse on Attic style. While admiration for Attic oratory is attested throughout the Hellenistic period, 24 we can safely say that it was the vehement controversy between Calvus’ movement and Cicero, immortalized in the latter’s rhetorical treatises of 46 BC, that marked the definitive breakthrough of Atticism into the rhetorical-stylistic discourse in Rome. Subsequently, both Romans and Greeks in the city took up this hot-button issue, appropriating the concept of Attic style and redefining it to suit their own purposes. All in all, we should unequivocally discard the false notion that the ancient students of Attic style can be linked together or separated from each other on the basis of their subscription to any kind of fixed stylistic program. Rather, ‘Attic’ was a versatile term that authors could employ to assert the superiority of their own work versus the ineptitude of their (‘Asian’) opponents. Therefore, it would be best, to quote Tim Whitmarsh, ‘to consider Atticism to have been an ever-negotiable concept, malleable according to the predilections and ambitions of the writer in question.’ 25 Rather than making conjectures about the history of Atticism by establishing links between Atticists through the centuries, I will review the Atticizing views of Calvus, Cicero and Dionysius in their contemporary contexts, as part of the aesthetic, cultural and political programs of the respective authors.

5.3 Attic Style and the Athenian Cultural Repertoire Before I move on to the various conceptions of Atticism in Rome, I will first establish the common set of parameters on the basis of which these views were articulated. One thing that Greek and Roman Atticists conspicuously have in common is their classicizing perspective: they invariably hark back to the Athenian prose models of the classical period, whilst rejecting the so-called ‘Asian’ authors of the subsequent Hellenistic era. 26 In addition, the

Φρυφῶν) and Demonstration that Every Word of Elegant Language has been Spoken (Ἀπόδειξις τοῦ εἰρῆσθαι πᾶσαν λέξιν καλλιρρημοσύνης ). Cf. section 1.5 n. 101 above. The Atticists of the Second Sophistic pay even more attention to linguistic matters: see Swain (1996) 20. O’Sullivan (2015), however, rightly notes that so- called ‘rhetorical’ and ‘linguistic’ Atticism should not be strictly separated. On the claim of Nassal (1910) 9–10 that Caecilius is the common source for the rhetorical and critical theories of Cicero and Dionysus, thus representing a link between the views of the two authors, see section 1.2 n. 12 above. 24 See the discussion of the Hellenistic antecedents of Atticism in Radermacher (1899a). In addition, we already saw that Agatharchides (second century BC) admired Attic oratory: see section 2.4.1 above. Note that Cic. Brut. 286 and Orat. 226 presents Hegesias (fl. ca. 300 BC) as an imitator of Lysias: cf. section 2.2 n. 28. 25 Whitmarsh (1998) articulates this view in response to Hidber (1996) 37–44: cf. n. 9 and 20 above. 26 On classicism and its inherent tripartite division of history, see section 1.6, esp. n. 114 above.

203

CHAPTER FIVE extant descriptions of Attic style, despite the overwhelming diversity between them, draw on a shared repertoire of attributes and virtues: they invoke the notions that are traditionally associated with the culture and politics of classical Athens. In this section, I will cursorily examine these notions, on which the discourse on Attic style in Rome is built. As we have seen, it is a much-cited adage in ancient stylistic theory that an author’s style reveals a lot about his character: style, indeed, is the man himself. 27 It naturally follows that Attic style should reproduce the national character of classical Athens. 28 In his speech in defense of Flaccus, Cicero offers us a glimpse of the prevailing reputation of the ancient polis in Rome: ‘There are present men from Athens, where reportedly civilization, learning, religion, agriculture, justice and laws were born and spread thence into every land. Tradition relates that even the gods competed for the possession of their city, so beautiful was it. It is so ancient that it is thought to have produced its own citizens, and the same soil is said to be their mother, their nurse and their home. Its prestige is so great that the present enfeebled and shattered renown of Greece is sustained by the reputation of this city.’29 This laudatory account of Athenian culture stands in a long and influential tradition, 30 that ultimately goes

27 See section 5.1 n. 4 above. 28 Cic. Orat. 25 explicitly connects a region’s literary style to the moral character of its inhabitants: ‘The eloquence of the orators has always been controlled by the good sense of the audience’ ( semper oratorum eloquentiae moderatrix fuit auditorum prudentia). Hence, Cicero explains the difference between Asian, Attic and Rhodian oratory by pointing out that Asian audiences are ‘the least civilized and least refined’ (minime politae minimeque elegantes ), whereas the judgments of the Athenians is ‘sound and discerning’ ( prudens sincerumque ) and the national character of Rhodians holds a middle between Asia and Athens. On Cicero’s reference to Rhodes as part of his apology for his rhetorical education, see section 1.6 above. 29 Cic. Flac. 62: Adsunt Athenienses, unde humanitas, doctrina, religio, fruges, iura, leges ortae atque in omnis terras distributae putantur; de quorum urbis possessione propter pulchritudinem etiam inter deos certamen fuisse proditum est; quae vetustate ea est ut ipsa ex sese suos civis genuisse ducatur, et eorum eadem terra parens, altrix, patria dicatur, auctoritate autem tanta est ut iam fractum prope ac debilitatum Graeciae nomen huius urbis laude nitatur. Cic. Flac. 65 also discusses the inhabitants of Caria, Phrygia and Mysia, addressing them not as Greek, but as Asian peoples. The three regions mentioned by Cicero often feature together in discussions of Asianism: see section 1.2 n. 36–37 above. 30 Cf. Lucr. 6.1–6: ‘It was Athens of illustrious name that first in former days spread abroad the corn-bearing crops amongst suffering mankind; Athens bestowed on them a new life and established laws; Athens first gave the sweet consolations of life, when she brought forth a man, endowed with such wisdom, who in past days poured forth all revelations from truth-telling lips’ ( primae frugiparos fetus mortalibus aegris / dididerunt quondam praeclaro nomine Athenae / et recreaverunt vitam legesque rogarunt, / et primae dederunt solacia dulcia vitae, / cum genuere virum tali cum corde repertum, / omnia veridico qui quondam ex ore profudit ). The ‘man’ ( vir ), introduced in the penultimate verse quoted above, is Epicurus.

204

TO BE ATTIC OR NOT TO BE ATTIC back to democratic Athens itself: Cicero’s words above resonate with the image of Athens that the Attic authors themselves had advertised. 31 It is this tradition that the critics and rhetoricians in Rome exploit in their discussions of Attic style.

Athens / Attic style Persia / Asian style

purity, integrity impurity, corruption

toughness, vigor softness, weakness

wisdom, sanity ignorance, insanity

freedom, democracy slavery, monarchy

moderation, restraint excess, extravagance

masculinity effeminacy

Table 9: typical attributes associated with Atticism and Asianism

Thus, our Atticists in fact anchor their diverse conceptions of Attic style in the patriotic rhetoric of Thucydides, Plato, Isocrates, Demosthenes and the other authors from the Attic canon that they admire. By appealing to the old Athenian stereotypes, authors can convincingly present their own stylistic theories as truly Attic. What, then, are the essential Athenian values that they take away from the age-old discourse? The table above presents several virtues and corresponding vices that, in Classical Attic literature, are associated with Athenian and Asian culture respectively: as we will see, these same notions are invoked in discussions about Attic and Asian style in the first century BC. 32 The list is not exhaustive, but merely aims to give an overview of the most salient recurring concepts, which the Greek and Roman Atticists in Rome could exploit to articulate their various interpretations of Attic style. The style of Asia, conversely, is typically presented as a foil to Atticism: the Asian vices are diametrically opposed to the Attic virtues. The biased description of Asian style goes back to the bashing of Persian culture in such Classical works as Aeschylus’ Persians ,

31 Isoc. Paneg. 28–39, for instance, claims that the Athenians invented religious rites, agriculture, laws and justice. Cf. ibid. 25 with Cic. Flac. 62 (esp. parens, altrix, patria ): ‘For we alone of all the Greeks have the right to call our city at once nurse and fatherland and mother’ ( μόνοις γὰρ ἡμῖν τῶν Ἑλλήνων τὴν αὐτὴν τροφὸν καὶ πατρίδα καὶ μητέρα καλέσαι προσήκει ). For Isocrates’ influence on Cicero, see Hubbell (1913) 16–40. 32 The ancient sources for the attributes in the list are discussed in the notes accompanying the text below. Forsdyke (2001) shows that the depictions of Greece and Persia from the fifth century BC are built on articulations of Athenian democratic ideology.

205

CHAPTER FIVE

Herodotus’ Histories and Isocrates’ Panegyricus . The denotations and connotations of Asia that these authors present accord well with the discourse on Asianism in Rome, and indeed with the cliché-ridden representation of the East throughout the history of Western Europe, which Edward Said has famously dubbed ‘Orientalism’. 33 The first item listed in the table above is concerned with autochthony: the Athenians often stress the purity of their race, claiming that they are indigenous to the soil of Attica. Isocrates, for example, boasts: ‘We did not become dwellers of this land by driving others out of it, nor by finding it uninhabited, nor as a mixed horde composed of many races, but we are of a lineage so noble and so pure that throughout our history we have continued in possession of the very land which gave us birth.’ 34 Atticists in Rome apply these notions of purity and autochthony to style: in his famous series of oratorical allegories, Dionysius, for one, describes the Attic muse as ‘old and indigenous’ (ἀρχαία καὶ αὐτόχθων ), while the Asian harlot is referred to as an ‘upstart who arrived yesterday or the day before’ (ἐχθὲς καὶ πρῴην ἀφικομένη ). 35 Likewise, Cicero claims that Attic eloquence remained ‘uncorrupted’ (incorrupta ), until it was exported abroad, especially to Asia, where it was inevitably stained by foreign ways. 36 The Atticists, then, advocate a return to the unadulterated language of Athens by stripping away barbarisms, solecisms and other alien elements. 37 It is important,

33 Said (1978), esp. 10–11: ‘Orientalism can accommodate Aeschylus, say, and Victor Hugo, Dante and Karl Marx.’ Much has been written on the stereotypical ways in which the Greeks depicted the ‘barbarian’ or the ‘other’ as means to articulate their own identities: see esp. Hartog (1980) on the representation of others in Herodotus, Hall (1989) on the ‘invention of the barbarian’ in Greek tragedy, and Cartledge (1993), who offers a description of Greek identity through a series of antitheses, including ‘Greek v. barbarian’. More recently, Gruen (2011), esp. 351–358, has problematized the emphasis on otherness that has dominated the recent scholarly debate, showing that the Greeks ‘postulated links with, adaptation to, and even incorporation of the alien’. 34 Isoc. Paneg. 24: Ταύτην γὰρ οἰκοῦμεν οὐχ ἑτέρους ἐκβαλόντες οὑδ’ ἐρήμην καταλαβόντες οὐδ’ ἐκ πολλῶν ἐθνῶν μιγάδες συλλεγέντες , ἀλλ ’ οὕτω καλῶς καὶ γνησίως γεγόναμεν , ὥστ ’ ἐξ ἧσπερ ἔφυμεν , ταύτην ἔχοντες ἅπαντα τὸν χρόνον διατελοῦμεν . Cf. also e.g. Thuc. 1.2.5, Eur. Ion 589, Ar. Vesp. 1076, Isoc. Pac. 49, Panath. 124–125. 35 Dion. Hal. Ant. orat. 1.6–7. This passage is quoted and discussed in section 5.4 below as one of three metaphors representing the fluidity of Atticism in Late-Republican and Augustan Rome. 36 For the uncorrupted nature of Attic style, see e.g. Cic. Brut. 51, a passage quoted and discussed in section 1.6 above. Cf. also Quint. Inst. orat. 12.10.19, who reproduces the view of the grammarian Santra (a contemporary of Cicero) that Attic style was corrupted through contact with foreign speech: cf. section 1.6 n. 129 above. 37 Through its emphasis on purity, Atticism is linked to Hellenism ( ἑλληνισμός ) and Latinity ( Latinitas ): see Porter (2006) 34–39. See also Dion. Hal. Lys. 2.1: Lysias is considered ‘pure’ ( καθαρός ) and therefore the ‘perfect model of the Attic dialect’ ( Ἀττικῆς γλώττης ἄριστος κανών ). Cf. Cic. Opt. gen. 7 on the ‘uncorrupted

206

TO BE ATTIC OR NOT TO BE ATTIC however, not to overdo it: according to Cicero, Calvus not only cleansed his speech from Asian errors, but also from its essential vigor. 38 Next, there seems to be a direct link between Attic style and the geography and ecology of Attica. The Athenians themselves stressed the ill-favored conditions of their land: the summers are dry and hot, and the soil is generally described as ‘thin’ (λεπτός ). 39 According to Plato, the landscape resembles the ‘skeleton of a sick man, all the fat and soft earth having wasted away, and only the bare framework of the land remaining.’ 40 This ecological discourse may be connected to Calvus’ interpretation of Atticism: his speech is described by his opponents as ‘dry’ (aridus, siccus ) and ‘thin’ (exilis, tenuis ), revealing its bare ‘bones’ (ossa ). 41 Thus, Calvus’ rugged oratory accurately reflects the landscape of Attica. Alternatively, Atticism could also be associated with the hard-bodied work of the farmers who have to cultivate the land: Cicero, accordingly, teaches us that Attic style should not merely attain ‘good health’ (valetudo ), but also ‘strength, muscles, blood and even, as it were, an attractive tan’ (vires, lacerti, sanguis, quaedam etiam suavitas coloris ). 42 The situation in Asia is wholly different: everything grows bigger, nature is tamer and richer, and health’ ( incorrupta sanitas ) of Attic style. Note also that the terms ‘solecism’ ( σολοικισμός ), after the town of Soli in Asia Minor, and ‘barbarism’ ( βαρβαρισμός ) are readily associated with Asianism. 38 See e.g. Cic. Brut. 283: ‘From excessive self-examination and fear of admitting error he lost true vitality’ (nimium inquirens in se atque ipse sese observans metuensque ne vitiosum colligerat, etiam verum sanguinem deperdebat ). For the lack of vigor in the oratory of Calvus and his movement, see Cicero’s critique in section 3.4 above and section 5.6.1 below. 39 See Thuc. 1.2.5, Pl. Criti. 111b–c, Men. Dys. 3, Str. Geogr. 9.1.8, Plut. Sol. 22, Lucian. Tim. 31. Redressing the Athenian descriptions of their own soil, Garnsey (1988) 95–96 shows that the climate in Attica is well-suited to the cultivation of barley and olives: cf. Theophr. Hist. pl. 8.8.2. Connors (1997) 69–72 shows that ‘natural connections between landscape and language’ underly the stylistic discourse in Rome: ‘The perceived differences between Attic and Asian rhetorical styles correspond exactly to perceived differences between the landscapes of Attica and Asia and the resulting physiognomic and psychological contrasts attributed to their inhabitants.’ 40 Pl. Criti. 111b: Οἷον νοσήσαντος σώματος ὀστᾶ, περιερρυηκυίας τῆς γῆς ὅση πίειρα καὶ μαλακή , τοῦ λεπτοῦ σώματος τῆς χώρας μόνου λειφθέντος . 41 For the meagerness and aridity of Calvus’ style, see esp. Cic. Brut. 68, 284–5, Tac. Dial. 21 and section 5.6.1. 42 Cic. Opt. gen. 8. In addition, Cic. Brut. 25 extols the ‘soundness’ ( salubritas ) and ‘health’ ( sanitas ) of Attic style. Note that Calvus and his followers also claimed that their meager style was ‘healthy’ ( sanus ) and ‘wholesome’ ( integer ): see Cic. Brut. 284. Cf. Dion. Hal. Dem. 32.1, who compares Dem. De cor. to ‘bodies developed by hard work in the sunlight’ ( ἐν ἡλίῳ πόνοις τεθραμμένα σώματα ), while Pl. Menex. is said to resemble bodies ‘that pursue a life of ease in the shade’ ( σκιὰς καὶ ῥᾳστώνας διώκοντα ). The Athenians were particularly proud of their agricultural feats: see, e.g., Pl. Menex. 237e–238a, Isoc. Paneg. 28 and Cic. Flac. 62.

207

CHAPTER FIVE the conditions are mild and unvarying. 43 Consequently, Asian style is naturally exuberant and characterized by the softness and idleness of those who live under the mellow sun of Asia Minor. 44 According to Cicero, a region’s weather conditions may also influence the intelligence of its population: hence, he asserts that the ‘thin climate’ ( tenue caelum ) of Attica rendered the Athenians ‘more intelligent’ ( acutiores ) than their neighbors. 45 Hence, virtues such as wisdom, sanity and cleverness are generally associated with both Athens and Attic style. The city was, of course, famous for its philosophers and sophists: the Athenians advertised their home as the ‘school of Greece’ ( Ἑλλάδος παίδευσις ) and the ‘sanctuary of wisdom’ (πρυτάνειον τῆς σοφίας ). 46 Likewise, Calvus and his followers considered themselves, as cultivators of Attic style, ‘wise men’ ( prudentes ). 47 In On the Ancient Orators , moreover, Dionysius describes Attic oratory as ‘philosophic’ ( φιλόσοφος ), while he presents its Asian counterpart, conversely, as ‘insane’ ( ἄφρων ), ‘ignorant’ ( ἀμαθής ), ‘brainless’ ( ἀνόητος ) and ‘not partaking in either philosophy or any other aspect of liberal education’ ( οὔτε φιλοσοφίας οὔτε ἄλλου παιδεύματος οὐδενὸς μετειληφυῖα ἐλευθερίου ). 48 Cicero, likewise, compares the

43 Cf. Connors (1997) 70, who quotes Hippoc. Aer. 12, 16: ‘Everything grows much bigger and finer in Asia and the one region (i.e., Asia) is less wild than the other (i.e., Europe), while the character of the inhabitants is milder and less passionate’ ( πολὺ καλλίονα καὶ μέζονα πάντα ἐν Ἀσίῃ, ἥ τε χώρη τῆς χώρης ἡμερωτέρη καὶ τὰ ἤθεα τῶν ἀνθρώπων ἠπιώτερα καὶ εὐοργητότερα ). 44 For the influence of the Asian climate on the character of the Persians, see e.g. Isoc. Paneg. 132. Hippoc. Aer. 16 notes: ‘With regard to the lack of spirit and of courage among the inhabitants, the chief reason why Asiatics are less warlike and more gentle in character than Europeans is the uniformity of the seasons, which show no violent changes either toward heat or toward cold, but are equable’ ( περὶ δὲ τῆς ἀθυμίης τῶν ἀνθρώπων καὶ τῆς ἀνανδρείης , ὅτι ἀπολεμώτεροί εἰσι τῶν Εὐρωπαίων οἱ Ἀσιηνοὶ καὶ ἡμερώτεροι τὰ ἤθεα αἱ ὧραι αἴτιαι μάλιστα , οὐ μεγάλας τὰς μεταβολὰς ποιεύμεναι οὔτε ἐπὶ τὸ θερμὸν οὔτε ἐπὶ τὸ ψυχρόν, ἀλλὰ παραπλησίω). For the association between Asia and wealth, and between Athens and poverty, see e.g. Dion. Hal. Ant. orat. 1.4 and Cic. Opt. gen. 9 as well as Aesch. Pers. 827–831, Hdt. 7.83.2, Pl. Leg. 694e, 695c–696a, Xen. Cyr. 8.8.7, 8.8.9– 12, 8.8.15–19, 8.8.26. On Asian softness and femininity, see the remainder of the present section and section 5.6 below. 45 Cic. Fat. 7 gives us another example of this kind of geographical determination: he mentions that the climate at Thebes is ‘dense’ ( crassum ), which has made the Thebans ‘stout and sturdy’ ( pingues et valentes ). For ‘thin’ and ‘fat’ as descriptions of Attic and Asian style respectively, see section 3.2 above. Cf. also section 3.2 n. 21 and tables 1 and 2 above. The adjective ‘thin’ ( tenuis ) refers to the simple style in, e.g., Cic. Orat. 20, 29, 53. 46 See, e.g., Thuc. 2.41.1, Pl. Prot. 337d–e, Isoc. Paneg. 48–50, Antid. 296. 47 See, e.g., Cic. Orat. 236, with section 3.6 above. 48 Dion. Hal. Ant. orat. 1.3–7. Cf. Hdt. 8.86 on the intelligence of the Athenians and the foolishness of the Persians. Aesch. Pers. 374–402 also contrasts the good sense of the Greek forces to the thoughtless disorder of

208

TO BE ATTIC OR NOT TO BE ATTIC

Attic and Asian styles on the basis of the ‘good sense’ ( prudentia ) of the citizens of Athens and the Asian towns: he attributes the success of Attic oratory to the ‘sound and discerning judgment’ ( prudens sincerumque iudicium ) of the Athenians, while the regions of Asia, which are ‘the least sophisticated and refined’ ( minime politae minimeque elegantes ), have yielded an inferior type of oratory. 49 In addition, Atticism and Asianism appeal to two diametrically opposed political systems and social models. Athens, for one, was regarded as the paradigm of freedom, democracy and the rule of law, stimulating its citizens to be self-restrained ( σώφρονες ). 50 According to Cicero, it is only in a ‘well-established civic order’ ( bene constituta civitas ) like Periclean Athens that eloquence can flourish. 51 Moderation and restraint, likewise, are archetypal Atticist virtues: Calvus and his movement, for example, insist that Attic style be ‘bound down by laws’ ( devinctum legibus ) and not embellished ‘beyond measure’ ( supra modum ). 52 Asia, conversely, stands for monarchy, slavery and disorder: the Persians, who once ruled the area, are generally referred to as a ‘mob without discipline’ ( ὄχλος ἄτακτος ), whose ‘whole existence consists of insolence toward some, and servility toward others’ and who can be seen ‘falling on their knees before a mortal man, addressing him as a divinity’. 53 Accordingly, slavery, excess and impiety are hallmarks of Asianism. 54 It is no coincidence that Dionysius portrays the Attic muse as a good Athenian citizen, ‘free and moderate’ (ἐλευθέρα καὶ σώφρων ), whereas her Asian challenger behaves like a Persian king, as she the Persian army. Isoc. Paneg. 150–156, lastly, enumerates the faults in the education ( παίδευσις ) of the Persians, while he praises Athens, as we have seen in n. 46 above, as the ‘school’ ( παίδευσις ) of Greece. 49 Cic. Orat. 25: cf. n 28 above. 50 The virtues of democracy and the rule of law are extolled in e.g. Thuc. 2.37, Dem. Arist. 206, Timocr. 210. The importance and various meanings of the term σωφροσύνη in Athenian society are discussed, by North (1966), who thinks that the word’s original sense is intellectual prudence, and by Rademaker (2005), who lists eightteen uses of the word. Cf. also section 5.5.2 and 5.6 below on Dionysius’ and Calvus’ focus on restraint. 51 Cic. Brut. 45. Cf. section 5.5.1, where we will see that Cicero contrasts the democratic constitution of Periclean Athens to the political situation of the Roman Republic in 46 BC. 52 See Quint. Inst. orat. 12.10.15 and Tac. Dial. 18.4. Cf. section 3.5 above and section 5.5.2 below for Dionysius’ focus on the ‘mean’ ( μεσότης ), ‘appropriateness’ ( πρέπον ) and ‘due measure’ ( μέτρον ). 53 Isoc. Paneg. 150–151: Ἅπαντα δὲ τὸν χρόνον διάγουσιν εἰς μὲν τοὺς ὑβρίζοντες τοῖς δὲ δουλεύοντες . Ibid.: Θνητὸν μὲν ἄνδρα προσκυνοῦντες καὶ δαίμονα προσαγορεύοντες . For the depictions of Greeks as free and well- disciplined, and of Persians as servile and weak-willed, see also Aesch. Pers. 230–245, Hdt. 5.78, 7.102–104, Pl. Leg. 693d–694a, 697c–e, and Xen. Cyr. 8.8.5. 54 See, e.g., Cic. Orat. 230: Asian authors are ‘slaves to rhythm’ ( numero servientes ). Quint. Inst. orat. 12.10.17 associates Asianism explicitly with lack of restraint: it is ‘bombastic’ ( tumidior ) and ‘redundant’ ( redundans ).

209

CHAPTER FIVE

‘claims control of the whole estate, treating the other like dirt and keeping her in a state of terror’. 55 The vocabulary of gender, which permeates the Atticist discourse, lastly, is also borrowed from the Classical testimonies about Athenian and Persian culture from the fifth and fourth centuries BC. The Athenian authors of that era enthusiastically asserted their ‘courage’, or ‘virility’ ( ἀνδρεία ), whereas they depicted Persian men as soft and effeminate, having no stamina for warfare at all. 56 Hence, it was an effective strategy for our Atticists to present themselves as true men, whilst challenging the manhood of their opponents. 57 Dionysius, for example, condemns Hegesias, the paradigm of Asianism, as effeminate: ‘His words are likely to be uttered only by women or emasculated men, and not even by them in earnest, but in the spirit of mockery and ridicule.’ 58 Similarly, Calvus and like-minded Atticists claim that Cicero is ‘softer than a man’ (viro mollior ), taunting him for being ‘emasculated’ (enervis, elumbis ). 59 Our Latin sources also use a variety of terms that refer less explicitly to femininity, such as ‘soft’ (mollis ), ‘loose’ (solutus ) and ‘broken’ (fractus ): they present Attic style, by contrast, as ‘whole’ (integer ). 60 Calvus more than any other Atticist stresses his own masculinity, accusing not only his oratorical opponents (esp. Cicero and Hortensius), but also his political foes (esp. Caesar and Pompey) of effeminate behavior (section 5.6). As we will see in the remainder of this chapter, the Athenian virtues that I have discussed in the present section provide the Greek and Roman Atticists in Rome with a

55 Dion. Hal. Ant. orat. 1.5: Πάσης ἀξιοῖ τῆς οὐσίας ἄρχειν , σκυβαλίζουσα καὶ δεδιττομένη τὴν ἑτέραν . The behavior of Dionysius’ Asian harlot resembles the reign of terror of the Persian king, as described in Isoc. Paneg. 151: the orator portrays the Persians as ‘keeping their souls in a state of abject fear, parading themselves at the door of the royal palace, prostrating themselves, and in every way schooling themselves to humility of spirit, addressing him as a divinity’ (τὰς δὲ ψυχὰς ταπεινὰς καὶ περιδεεῖς ἔχοντες , ἐξεταζόμενοι πρὸς αὐτοῖς τοῖς βασιλείοις καὶ προκαλινδούμενοι καὶ πάντα τρόπον μικρὸν φρονεῖν μελετῶντες , θνητὸν ἄνδρα προσκυνοῦντες καὶ δαίμονα προσαγορεύοντες ). 56 Roisman (2005) discusses the importance of manhood (as well as the flexibility of that notion) in the rhetoric of the Attic orators. On the effeminacy of the Persians, see already Aesch. Pers. 845–851 with Hall (1989) 117– 120, and Hdt. 1.155.4. 57 Richlin (1997) 106–107 discusses the role of gender in the extant Latin articulations of Atticism and Asianism. Connolly (2007b) explains how Roman orators and rhetoricians could draw on the language of gender so as to assert their own superiority over their opponents. 58 Dion. Hal. Comp. 18.28: Ὠς δὲ ὁ Μάγνης εἴρηκεν , ὑπὸ γυναικῶν ἢ κατεαγότων ἀνθρώπων λέγοιτ ’ ἂν καὶ οὐδὲ τούτων μετὰ σπουδῆς, ἀλλ ' ἐπὶ χλευασμῷ καὶ καταγέλωτι . Cf. section 2.4.3 above. 59 See Quint. Inst. orat. 12.10.12 and Tac. Dial. 18.4, quoted and discussed in section 5.6.1 below. 60 See esp. section 5.6 below on Calvus’ conception of Attic style as sober, pure and masculine.

210

TO BE ATTIC OR NOT TO BE ATTIC common, though by no means uniform, conceptual framework: Calvus, Cicero, Dionysius and their colleagues each make a different selection from this shared repertoire of Athenian cultural products to suit their various purposes. Specifically, the authors tailor their conceptions of Athens and Attic style not only to their respective stylistic programs and aesthetic tastes, but also to their political views, demarcating their own position as citizens of Late-Republican or Augustan Rome respectively.

5.4 Atticism Personified: the Various Attic Muses of Rome Metaphors can offer quick, clear and pleasant insights into difficult matters. This section briefly discusses three metaphors that illustrate the Atticist stylistic theories of Calvus, Cicero and Dionysius. In each of these cases, the sources compare eloquence to a fair and chaste lady: Calvus’ lady is an ‘unadorned woman’ ( mulier inornata ), Cicero’s is an ‘orphaned virgin’ ( virgo orba ) and Dionysius’ is a ‘freeborn and faithful wife’ ( ἐλευθέρα καὶ σώφρων γαμετή ). 61 As we will see, these three Attic muses each stand for a different conception of Atticism (sections 5.5 and 5.6). Thus, the various feminine personifications illustrate at a glance the versatility of the Atticizing stylistic discourse in Rome. For Calvus’ muse, first, we should turn to his oratorical adversary: when Cicero discusses the wholesome qualities of the sober, unadorned orator, ‘whom some deem to be the only true Attic orator’, 62 he introduces the image of a naturally beautiful woman whose figure is perfectly in line with Calvus’ views about oratorical style.63

Nam ut mulieres pulchriores esse dicuntur nonnullae inornatae quas id ipsum deceat, sic haec subtilis oratio etiam incompta delecta; fit enim quiddam in utroque, quo sit venustius sed non ut appareat. Tum removebitur omnis insignis ornatus quasi

61 The fact that most ancient allegorical personifications are feminine has baffled modern scholars: Paxson (1998) and Stafford (1998) show that this predilection for femininity should not simply be attributed to grammatical formalism, that is, the idea that the personfication of rhetoric ( ῥητορικὴ τέχνη , eloquentia ) is female because of the grammatical gender of the noun. Paxson argues that ‘personification as a concept could be thought of as having the gendered qualities of the feminine’, as figurative language was associated with ‘ornamentation, seduction, excess’. According to Stafford, the form of the female body, as the object of men’s desire, ‘conveys the desirability of the abstract values’ that it embodies. Leidl (2003) collects and discusses various female personifications of eloquence from the works of Cicero, Tacitus, Dionysius and Porphyry. 62 Cic. Orat. 75: Quem solum quidam vocant Atticum. The word ‘some’ (quidam ) refers to the self-styled Attic orators in Rome: see section 3.4 above for Cicero’s attack on their conception of the plain style. 63 Cic. Orat. 78–79.

211

CHAPTER FIVE

margaritarum, ne calamistri quidem adhibebuntur. Fucati vero medicamenta candoris et ruboris omnia repellentur: elegantia modo et munditia remanebit.

Just as some women are said to be handsomer when unadorned—this very lack of ornament becomes them—so this plain style gives pleasure even when unembellished: there is something in both cases which lends greater charm, but without showing itself. Also, all noticeable ornaments, pearls as it were, will be excluded; not even curling- irons will be used; all cosmetics, artificial white and red, will be rejected, only elegance and neatness will remain.

Without referring to his opponent explicitly, Cicero personifies Calvus’ Atticizing oratory: the image of the unembellished woman tallies neatly with Calvus’ program of purity and restraint. Moreover, as we will see, the orator attacks Cicero for the degenerate effeminacy of his bombastic speeches: in the terms of the allegory, Calvus takes offense at the use of eye liner, rouge or lipstick to adorn one’s oratory (section 5.6.1). The keywords ‘elegance’ (elegantia ) and ‘neatness’ ( munditia ) that Cicero uses to describe the unadorned woman, are both closely connected to the plain type of style, which Calvus holds so dear. 64 In addition, Roman rhetoricians often associate the use of lavish ornamentation with luxurious make-up for women: according to Suetonius, for example, the emperor Augustus compared the frivolous style of his friend Maecenas to ‘perfume-dripping locks’ ( myrobrechis cincinni ). 65 Likewise, the younger Seneca argues that style should not be ‘trimmed’ ( circumtonsa ), ‘dyed’ (fucata ) or ‘groomed’ ( manu facta ): the Stoic philosopher concludes that such extravagance does not constitute a ‘masculine ornament’ ( ornamentum virile ). 66 As we will see, this purist approach lies at the core of Calvus’ Atticist program: in his view, ostentation and

64 The virtue of ‘elegance’ ( elegantia ) is typically associated with the simple style. Rhet. Her. 4.17, for one, defines the term as ‘that which makes each and every topic seem to be expressed with purity and perspicuity’ (quae facit ut locus unus quisque pure et aperte dici videatur ), dividing it into the subsections of ‘Latinity’ (Latinitas ) and ‘clarity’ ( explanatio ), two typical attributes of simple rhetoric. López Moreda (2003) reviews the various uses of elegantia as a grammatical, rhetorical and critical term: she notes that the concept is closely connected to the notion of Latinity also bearing on the ‘balance between the mean and the redundant, overloaded style’ (‘equilibro entre el estilo bajo y el redundante y sobrecargado’). As for the word ‘neatness’ (munditia ), Quint. Inst. orat. 8.3.87 associates it with ‘plain and unaffected simplicity’ (ἀφέλεια simplex et inadfectata ). Cf. Gell. NA 1.23.1 and 10.3.4 on the ‘neatness of words’ ( munditia verborum, mundities orationis ). 65 Suet. Aug. 86.2. See Sen. Ep. 114.4–8 for the allegedly effeminate style of Maecenas. Cf. section 5.6.1 below. 66 Sen. Ep. 115.2–3. Cf. section 4.5.1 n. 89 above and section 5.6.1 n. 140 below.

212

TO BE ATTIC OR NOT TO BE ATTIC extravagance are corrupting, feminizing and Asianizing elements that have no place in Rome, neither in Attic oratory nor in the public life of the Forum. Cicero, as we have seen, thinks that Calvus’ approach to Attic style is misguided (section 3.4). In his Brutus , he presents an alternative portrait of Lady Eloquence, which illustrates his own conception of Atticism. The passage below concludes Cicero’s eulogy of the great orator Hortensius, who had died a few years earlier, in 50 BC, just before the outbreak of the civil war between the armies of Caesar and Pompey. 67

Nos autem, Brute, quoniam post Hortensi clarissimi oratoris mortem orbae eloquentiae quasi tutores relicti sumus, domi teneamus eam saeptam liberali custodia et hos ignotos atque impudentis procos repudiemus tueamurque ut adultam virginem caste et ab amatorum impetu, quantum possumus, prohibeamus.

As for us, Brutus, since with the death of Hortensius, the brilliant orator, we are left to be the guardians of orphaned eloquence, let us keep her within our own walls, protected by a private custody. Let us repel the pretensions of these ignoble and impudent suitors, and guard her purity, like that of a virgin grown to womanhood, and, so far as we can, shield her from the advances of rash admirers.

In 46 BC, two years into Caesar’s dictatorship, Cicero represents ‘oratory’ ( eloquentia ) as an orphan: after the death of Hortensius, she now requires new ‘legal guardians’ ( tutores ) to keep her at home in ‘private custody’ ( liberalis custodia ), where she will be safe from malevolent ‘suitors’ ( proci, amatores ), who are out to corrupt her. 68 Like Calvus, then, Cicero wants to preserve the purity of eloquence by warding off perverting elements. Yet, unlike Calvus, Cicero is not so much worried about the baleful influence of effeminate Asianists; if anything, he shows himself quite lenient to the so-called Asianist oratory of his rival Hortensius

67 Cic. Brut. 330. On Hortensius, in whose hands Cic. De or. 3.228-230 placed his hope for the future of eloquence, see esp. section 5.6.1 below. 68 Leidl (2003) 35–38 notes that orphanage is a recurring metaphor in Cicero’s works: see, e.g., Red. sen. 4, Quir. 11, De or. 3.1, Fam. 3.11.3, Leg. 3.3.9 and Flac. 54. Cicero usually uses the image to illustrate the deplorable state of the Republic, which he considers deprived of the necessary support in a time of turmoil, like a child bereft of its mother. According to Leidl, therefore, the passage quoted here is ‘merging the literary and political judgment through the use of metaphor’. In one of the mss (L), ‘armed men’ ( armatorum ) is transmitted instead of ‘lovers’ ( amatorum ): the error may have been elicited by the fact that Cicero refers to the detrimental effects of civil war and Caesar’s violent reign.

213

CHAPTER FIVE

(sections 1.6 and 5.6.1). Rather, he fears the freedom-restricting policies of Caesar and his cronies: now that Lady Eloquence has become a ‘grown-up woman’ ( adulta virgo ), these ‘ignoble and impudent suitors’ ( ignoti et impudentes proci ) will launch an ‘attack’ ( impetus ) on her chastity. 69 According to Cicero, Caesar’s dictatorship has effectively destroyed the freedom of the republic, making it extremely hard for the orators in Rome to practice their art (section 5.5.1). Therefore, Cicero’s Attic muse is a damsel in distress, in need of a strong, passionate defender, that is, not a subdued Lysias or an emaciated Calvus, but rather a vigorous Demosthenes or a vehement Cicero. What, finally, is the situation in Augustan Rome? According to Dionysius, the new autocratic regime did not bring about the death of eloquence, as Cicero had feared. On the contrary, Dionysius praises the rulers of Rome for restoring the Attic muse to her former glory (section 5.5.2). In the introductory essay to On the Ancient Orators , the Greek critic not only personifies Attic oratory, but also its Asian counterpart. 70

Ὥσπερ γὰρ ἐν ἐκείναις ἡ μὲν ἐλευθέρα καὶ σώφρων γαμετὴ κάθηται μηδενὸς οὖσα τῶν αὑτῆς κυρία , ἑταίρα δέ τις ἄφρων ἐπ’ ὀλέθρῳ τοῦ βίου παροῦσα πάσης ἀξιοῖ τῆς οὐσίας ἄρχειν , σκυβαλίζουσα καὶ δεδιττομένη τὴν ἑτέραν· τὸν αὐτὸν τρόπον ἐν πάσῃ πόλει καὶ οὐδεμιᾶς ἧττον ἐν ταῖς εὐπαιδεύτοις (τουτὶ γὰρ ἁπάντων τῶν κακῶν ἔσχατον ) ἡ μὲν Ἀττικὴ μοῦσα καὶ ἀρχαία καὶ αὐτόχθων ἄτιμον εἰλήφει σχῆμα , τῶν ἑαυτῆς ἐκπεσοῦσα ἀγαθῶν, ἡ δὲ ἔκ τινων βαράθρων τῆς Ἀσίας ἐχθὲς καὶ πρῴην ἀφικομένη , Μυσὴ ἢ Φρυγία τις ἢ Καρικόν τι κακόν , Ἑλληνίδας ἠξίου διοικεῖν πόλεις ἀπελάσασα τῶν κοινῶν τὴν ἑτέραν , ἡ ἀμαθὴς τὴν φιλόσοφον , ἡ μαινομένη τὴν σώφρονα .

Just as in such households there sits the lawful wife, freeborn and faithful, but with no authority over her domain, while an insensate concubine, bent on destroying her livelihood, claims control of the whole estate, treating the other like dirt and keeping her in a state of terror; so in every city, and in the highly civilized ones as much as any

69 Note that Cicero depicts his Lady Eloquence as ‘grown-up’ ( adulta ), implying that Roman oratory has for a long time been immature, until it finally reached maturity during the lifetime (and because of the efforts) of Hortensius and Cicero. Cf. Stroup (2003) on Cicero’s personification of eloquence as a grown-up woman. 70 Dion. Hal. Ant. orat. 1.5–7. The word ἐκείναις in the first line refers to ‘households’ (οἰκίαις ) in Ant. orat. 1.5. The passage has been amply discussed as an allegorical defense of classicism: see esp. Hidber (1996) 25–30, Leidl (2003) and De Jonge (2014a); cf. n. 71–72 below. For Dionysius’ praise of Rome, which will be discussed in section 5.5.2 below, see Ant. orat. 3.1–3.

214

TO BE ATTIC OR NOT TO BE ATTIC

(which was the final indignity) the ancient and indigenous Attic Muse, deprived of her possessions, had lost her civic rank, while her antagonist, an upstart that had arrived only yesterday or the day before from some Asian death-hole, a Mysian or Phrygian or a Carian creature, claimed the right to rule over Greek cities, expelling her rival from public life. Thus, the ignorant woman drove out the philosophical one, the mad woman the sound one.

Both Cicero and Dionysius portray eloquence as a chaste woman who is in danger of being driven out of her home. 71 While Cicero, however, depicts her as an unmarried woman who needs protection from depraved (male) suitors, Dionysius presents her as a ‘lawful wife’ (γαμετή ), who is outrageously expelled from her home by a (female) ‘concubine’ ( ἑταίρα ). Unlike Cicero, Dionysius feels no need to undertake rash action to save his Attic muse: in On the Ancient Orators , he emphatically celebrates the victory of Atticism and the concomitant virtues of Classical Athens. 72 the critic twice mentions that his Attic muse is σώφρων , a word that not only evokes the prudence of a loyal Athenian housewife, but also such notions as self- control and moderation, on which Dionysius builds his conception of the middle style (section 3.5). 73 The reasonable, wise and thoughtful Athenian lady is an apt personification for Dionysius’ Atticism, which is infused with such notions as moderation, justice and piety. At

71 Leidl (2003) 43–47 notices that both Cicero and Dionysius draw a contrast between ‘in’ and ‘out’. Cicero’s orphan, for one, needs to be kept ‘at home’ ( domi ), while her suitors need to be kept out. Likewise, Dionysius depicts the Attic muse as a house-wife, who is expelled from her house by an external force. See also De Jonge (2014a) 395–396. 72 Cf. section 5.4 table 9 above: esp. the notions of purity, freedom, wisdom and moderation are relevant for Dionysius’ discussion of Attic style. The allegory of Ant. orat. 1.5–7 draws on several themes that can be found in earlier sources. Hidber (1996) 25–30, for example, points out similarities between Dionysius’ ‘philosophical’ (φιλόσοφος ) Attic muse, and Isocrates’ conception of rhetoric as a kind of ‘philosophy’ ( φιλοσοφία ): cf. Livingstone (2007). According to Wiater (2011) 92–100, Dionysius uses his allegory to assert the Greek superiority over other peoples, presenting the battle between Atticism and Asianism as an installment in the ‘prolonged, Greek struggle against the Barbarians’; cf. section 1.2 n. 26 above. De Jonge (2014a) 395 connects the passage to Prodicus’ story about ‘Hercules at the crossroads’, choosing between the female personification of ‘virtue’ ( ἀρετή ) over that of ‘vice’ ( κακία ): see Xen. Mem. 2.1.21. 73 The two translations of the Greek term σώφρων reflect the ‘polysemy’ of the word, as is described by Rademaker (2005). I have translated the first instance of σώφρων as ‘faithful’, since it is used specifically in the context of a woman’s role as a ‘lawful wife’ ( γαμετή ): cf. Rademaker (2005) 252–253. I rendered the second instance of σώφρων , however, as ‘sane’, as it stands in opposition to the word ‘mad’ ( μαινομένη ): cf. Rademaker (2005) 260–261.

215

CHAPTER FIVE the same time, the image of the devoted spouse bears some resemblance to the self- presentation and the politics of Augustus: the emperor emphasized his own self-restraint and humility, while he introduced marital laws like the Lex Iulia de Maritandis Ordinibus (18 BC) and the Lex Iulia de Adulteriis Coercendis (17 BC), severely punishing celibacy and infidelity (section 5.5.2). Calvus’ naturally beautiful woman, Cicero’s impressionable young lady, and Dionysius’ loyal wife each stand for a different interpretation of Attic style. In the following sections, I will take a closer look at the Atticizing perspectives of the three aforementioned authors, focusing on the intricate connections between their stylistic views on the one hand and their respective approaches to contemporary Roman politics and society on the other hand. In what way does the Atticism in the works of Calvus, Cicero and Dionysius contribute to their cultural, moral and political self-presentation in Rome? It is to this question that I will now turn.

5.5 The Politics of Stylistic Theory: Cicero and Dionysius on Attic Style The names of Calvus, Cicero and Dionysius can each be related to a specific stage in the transition from Republic to Empire. The record of Calvus’ activity is limited to the mid-50s of the first century BC, when Caesar, Pompey and Crassus had solidified their allegiance to each other in the so-called First Triumvirate (60–53 BC). Cicero composed his Brutus , Orator and De Optimo Genere Oratorum , when Caesar had fulfilled two years of his ten-year appointment as dictator of Rome, which would end prematurely two years later (48–44 BC). 74 Dionysius arrived in the city ‘right at the very time when Augustus Caesar put an end to the civil war’, viz., at the very beginning of Imperial Rome (ca. 30 BC).75 In the next two sections, we will see that the views on Attic style that the three authors bring to the fore are closely connected to their opinions about contemporary politics. Once again, I will demonstrate the importance of the ancient adage that the selection and the arrangement of words should reflect the thought and the beliefs of their speaker—literary style as the emblem of a man’s soul.76 The opinions on prose style of the oldest Atticist in our record, Calvus, not only tally with his political views but also with the information that we have about his poetry and about

74 On Cicero’s works of 46 BC and on the sources for Calvus’ oratory, see section 1.4 above. 75 Dion. Hal. Ant. Rom. 1.7.2: Ἅμα τῷ καταλυθῆναι τὸν ἐμφύλιον πόλεμον ὑπὸ τοῦ Σεβαστοῦ Καίσαρος . Cf. also section 1.1 n. 2 above. 76 Cf. section 5.1 n. 4 above.

216

TO BE ATTIC OR NOT TO BE ATTIC his personal life: therefore, the pervasiveness of Atticism in Calvus’ entire life merits a separate discussion (section 5.6). The present section will focus on the interplay between politics and stylistic theory in the rhetorical works of Cicero and in the critical treatises of Dionysius. In what way do these two men use their critical programs to demarcate their respective positions in the Roman political arena of their day?

5.5.1 Cicero’s Sublime Oratory and the Battle for the Republic In the last three years of his life (46–43 BC), when he had just turned sixty, Cicero suddenly increased his literary output: in this period, the former consul not only composed four rhetorical treatises, but he also added at least fourteen philosophical works to his impressive oeuvre.77 This conspicuous turn from the oratorical practice of the Forum to theorization in Tusculum is an obvious reaction to the current political situation in Rome: without mentioning Caesar by name, the ex-consul repeatedly notes that the rule of law and the power of eloquence had dwindled under the threat of the dictator’s weapons. In his Brutus , Cicero complains: ‘It is a source of deep pain to me that the state feels no need of those weapons of counsel, of insight, and of authority, which I had learned to handle and to rely upon, weapons which are the peculiar and proper resource of a leader in the commonwealth and of a civilized and law-abiding state.’ 78 Hence, Cicero saw no other viable course of action than to retreat largely from public life and turn to his literary studies, as he explains in his Orator : ‘Who would be hard or unfeeling enough to refuse me the favor of devoting myself to letters (litterae ), now that my forensic practice and my public career have fallen in ruins, rather than to idleness, which is impossible to me, or to grief, against which I put up a bold front?’ 79

77 Cicero’s rhetorical works of this period are Brut., Opt. gen. , Orat. and Top. ; his philosophical treatises are Parad., Hortensius, Acad. pr. (or, Lucullus ), Acad. post. (or, Varro ), Consolatio , Fin. , Tusc. , Nat. D. , Div. , Fat. , Sen., Amic. , De Gloria and Off. From the same period, we have three speeches ( Marcell. , Lig. , Deiot. ) in addition to the fourteen orations against Antony ( Phil. ). Cf. the overview in Steel (2013) 374–376. 78 Cic. Brut. 7: Equidem angor animo non consili, non ingeni, non auctoritatis armis egere rem publicam, quae didiceram tractare quibusque me assuefeceram quaeque erant propria cum praestantis in re publica viri tum bene moratae et bene constitutae civitatis. See also ibid. 21: ‘Devastation of the courts and the forum’ (iudiciorum vastitas et fori ). Cf. ibid. 22: ‘Eloquence has become mute’ ( eloquentia obmutuit ). On Cicero’s opaque criticism of Caesar, see Stroup (2003) 122 and Dugan (2012), who consider Brut. a good example of a ‘figured’ critique, which was common feature imperial literature. See Chiron (2001) 224–236 on ‘figured speech’ ( λόγος ἐσχηματισμένος ), cf. Demetr. Eloc. 287–95, Long. Subl. 44.3–4, Quint. Inst. orat. 9.2.65–99. 79 Cic. Orat. 148: Quis tamen se tam durum agrestemque praeberet, qui hanc mihi non daret veniam, ut cum meae forenses artes et actiones publicae concidissent, non me aut desidiae, quod facere possum, aut maestitiae,

217

CHAPTER FIVE

Cicero’s dedication to literature should not be thought to mark his complete withdrawal from politics. I do not agree with John Dugan’s conclusion that the author of Orator is ‘resigned to the loss of the republic’, as he ‘fills his days with social calls from both political allies and Caesarians alike, providing oratorical lessons, reading, writing, and engaging in the care of the body’. Likewise, I disagree with Dugan’s view that Cicero’s extensive discussion of prose rhythm ‘marks a distinctive aesthetic turn in Cicero’s treatment of oratory, away from oratory as a political instrument or as a product of historical development, and toward its status as an autonomous work of art’. 80 As a matter of fact, Cicero’s rhetorical works of 46 BC are not the musings of a detached pensionado: his discussions of oratorical style are imbued with a genuine political engagement. After the assassination of Caesar, Cicero emphasizes the political nature of his recent theoretical treatises: ‘It was in my books that I made my senatorial speeches and my forensic harangues, for I regarded philosophy as my substitute for a role in politics.’ 81 Hence, as Cicero’s cui resisto, potius quam litteris dederem. On Cicero’s apology for focusing on technical, seemingly unpractical, topics in Orat. 140–148, see section 4.2 n. 32–33 above. For Cicero’s motivations to write philosophy, see Baraz (2012) esp. 173–182, who finds three motivations to write philosophy in the prefaces to Cicero’s treatises—one political (‘to serve the state and its citizens’), one personal (‘to keep active even when he cannot directly take part in politics’), and one consolatory (‘to find solace after the death of his daughter’ in 45 BC). Baraz also notes that Cicero’s retreat from public life in the 40s BC is in two important respects more dramatic than his exile in the 50s BC: first, the exile is a personal sorrow, whereas under Caesar’s dictatorship he ‘is living through an exclusion that is shared with others; he is part of a class, in fact more than just a part: as one of the eldest consulars to survive the civil war, he is seen by some as the leader of this group’; secondly, in the 50s, he ‘saw the basic functioning of the state as still unchallenged’, while under Caesar ‘fundamental changes to the institutions of the state, embodied most dramatically in, but not limited to, the role of the dictator , resulted in a political landscape very different from the republican model’. 80 Dugan (2005) 253–258. According to Dugan, Orat. anticipates a feature that would become prominent in the literary culture of the empire, that is, the domestic teaching of oratory through delivering one’s own declamations and listening to those of one’s pupils. For Dugan’s influential readings of Brut. and Orat. , see also sections 1.2 and 1.4 above. 81 Cic. Div. 2.7: In libris enim sententiam dicebamus, contionabamur, philosophiam nobis pro rei publicae procuratione substitutam putabamus. Cic. Fam. 9.18.1–2 spells out Cicero’s political intentions in teaching such pupils as Dolabella and Aulus Hirtius, two supporters of Caesar: ‘From your letter I gathered that my plan meets with your approval: like Dionysius the tyrant, who is said to have opened a school at Corinth after his expulsion from Syracuse, I have set up as a schoolmaster, as it were, now that the courts are abolished and my forensic kingdom lost’ ( ex quibus intellexi probari tibi meum consilium, quod, ut Dionysius tyrannus, cum Syracusis pulsus esset, Corinthi dicitur ludum aperuisse, sic ego sublatis iudiciis, amisso regno forensi ludum quasi habere coeperim ).

218

TO BE ATTIC OR NOT TO BE ATTIC influence in the courts and the assemblies waned, he turned to philosophy and rhetorical theory as alternative strategies to assert his political authority. In what way do the stylistic discussions of Brutus, Orator and De Optimo Genere Oratorum express Cicero’s involvement in the political turmoil of the mid-40s BC? The flexible notion of Attic style provides Cicero with a twofold vehicle to combine stylistic theory and political engagement in his rhetorical treatises. First, Atticism allows him to stress the inextricable entanglement of good government and brilliant eloquence. According to Cicero, it is in the career of the orator-statesman Pericles—that is, after the Athenians had defeated Persia, and before they entered into a devastating struggle with Sparta—that oratory started to flourish for the first time in history: ‘This age therefore first produced at Athens an orator all but perfect. For the ambition to speak well does not arise when men are engaged in establishing government, nor occupied with the conduct of war, nor shackled and chained by the authority of kings. Upon peace and tranquility eloquence attends as their ally; it is, one may say, the offspring of well-established civic order.’ 82 As we saw before, Pericles is one of Cicero’s principal models for grand oratory—the type of style that he commends as the most powerful in winning cases (sections 2.3.2 and 3.3). Needless to say, however, the situation in Rome is far removed from the Periclean ideal: the city is tormented by constitutional crises, torn apart by civil wars, and consequently bereft of oratorical brilliance. 83

82 Cic. Brut. 45: Haec igitur aetas prima Athenis oratorem prope perfectum tulit. Nec enim in constituentibus rem publicam nec in bella gerentibus nec in impeditis ac regum dominatione devinctis nasci cupiditas dicendi solet. Pacis est comes otique socia et iam bene constitutae civitatis quasi alumna quaedam eloquentia. Venini (1976) shows that Cicero highlights Pericles’ outstanding statesmanship, ignoring his controversial ideas about radical democracy. Pericles’ famous funerary oration is one of Cicero’s (and Dionysius’) favorite speeches: see sections 2.3.2 n. 59 and 2.4.3 n. 147 above. Cic. Brut. 46 connects the earliest beginnings of rhetorical theory in Syracuse to the expulsion of the tyrant Thrasybulus: ‘Thus, Aristotle says that in Sicily, after the expulsion of tyrants, when after a long interval restitution of private property was sought by legal means, Corax and Tisias the Sicilians, with the acuteness and controversial habit of their people, first put together some theoretical precepts’ (itaque ait Aristoteles, cum sublatis in Sicilia tyrannis res privatae longo intervallo iudiciis repeterentur, tum primum, quod esset acuta illa gens et controversa natura, artem et praecepta Siculos Coracem et Tisiam conscripisse ). Cf. the next note on the ancient discourse about the political causes of corrupt eloquence. 83 The idea that tyranny inevitably leads to ‘corrupt eloquence’ ( corrupta eloquentia ), in the same way as democracy fosters oratorical genius, was a common trope in discussions about rhetoric and style in the first century AD. Quintilian wrote a (now lost) treatise On the Causes of Corrupt Eloquence (De causis corruptae eloquentiae ): cf. Quint. Inst. orat. 5.12.23, 6.Praef.3, with Brink (1982). Long. Subl. 44 cites an anonymous philosopher who expressed the view that literature can only prosper under the scepter of freedom and democracy, whereas the present imperial institutions impede the talents of aspiring authors. Additionally, Tac.

219

CHAPTER FIVE

Secondly, Cicero presents the stylistic grandeur of Attica not only as a nostalgic counterpart to the oppression of eloquence in Rome, but also as an antidote to the current disastrous state of affairs. 84 According to Cicero’s diagnosis, the city needs forceful, vehement orators in order to save the ‘orphaned Lady Eloquence’ ( eloquentia orba ) from her ‘ignoble and impudent suitors’ ( ignoti atque impudentes proci ), as we saw above (section 5.4). In addition to his other objections to the claims of Calvus and his fellow Atticists (section 3.4), this is an important reason for Cicero to reject the subdued style of his opponents: in his view, the current situation does not require a cautious and subtle orator, but rather one who relentlessly pulls out all the stops to win the day. 85 The fact that CIcero engages in a lengthy study of such stylistic intricacies as word arrangement, periodic structure and rhythmical cadences should not be taken as proof for his definitive withdrawal from forensic politics to his study room: on the contrary, it is only through these devices that, in his view, the republic can be saved. Therefore, the grand style in general and Demosthenes’ conception of it in particular are presented as the best models for Rome. Cicero especially praises the force of Demosthenes’ rhythm: ‘Those famous thunderbolts of his would not have sped with such vibrant power, if they had not been whirled onward by rhythm.’86 According to Cicero, it is through the exuberant rhythm and the bombastic periods of grand oratory that an orator can sway the minds of the masses on the Forum. 87

Dial. 36–42 ascribes to the tragedian Curiatius Maternus the opinion that eloquence can only flourish, if it is free from any kind of rule: hence, he argues that it was during the anarchy of the civil wars of Republican Rome that an excellent form of oratory could arise, which would become obsolete during the quiet peace of the Empire. For a good discussion of ancient discussions of corrupt eloquence, see Caplan (1970). Cf. also Kennedy (1972) 446– 464, 494–496, 515–526, who claims that the earliest treatment of oratorical decline appears in Long. Subl .; and Heldmann (1982) 300–308, who instead argues that Tac. Dial. was the first source on such decay. Yet, as we have seen in the present section, the topic already attracted the attention of Cicero: cf. Cic. Ver. 2.5.31 (70 BC) and Pis. 32 (55 BC). See, in contrast, Dionysius’ optimism about imperial oratory in Ant. orat. (section 5.5.2). 84 See also Cic. Brut. 53 for the view that in 509 BC the Tarquins could never have been expelled from Rome and the republican constitution could never have been installed, ‘if it were not for the persuasion of oratory’ ( nisi esset oratione persuasum ). 85 Cf. Cic. Opt. gen. 10 defending the application of the grand style in Mil. : see section 2.3.2 n. 60–63 above. According to Cicero, one of the distinctive qualities of Demosthenes’ oratory is its usefulness in life-and-death situations: see section 2.4 above. 86 Cic. Orat. 234: Cuius non tam vibrarent fulmina illa, nisi numeris contorta ferrentur. On Pericles’ and Demosthenes’ thunderbolts, see section 2.3.2, esp. n. 59, and section 3.3 above. 87 On Cicero’s view that the grand style speaks to the uneducated masses, while the simple style only pleases a minority of intellectuals, see section 3.6 above. On Demosthenes’ ability to mobilize Greece, see Cic. Brut. 289.

220

TO BE ATTIC OR NOT TO BE ATTIC

It has often been remarked that Cicero presents himself as a Roman Demosthenes. 88 Indeed, the Greek orator could be regarded as the last staunch defender of Greek freedom in the face of Philip’s tyranny, just as Cicero sees himself as Rome’s last bulwark against Caesar’s violent dictatorship. 89 In the autobiographical section of his Brutus , Cicero forges various striking connections between his own career and Demosthenes’ life, as is expounded by Caroline Bishop: ‘Both trained long and hard, individually and with teachers, to conquer natural disadvantages, both came off as overly forceful before maturing into more varied styles, and both embodied a philosophical rhetoric reminiscent of Plato.’ 90 It is probably no coincidence that Cicero singles out On the Crown as his favorite Demosthenic speech, as it sums up both the politician and the orator that Cicero aspires to be: it not only cleverly combines different stylistic registers, but it also dwells at length on the faults of tyranny, commending oratory as an effective strategy to expel tyrants such as Philip of Macedon.91 Positioning himself as a worthy successor to Demosthenes’ versatile anti-tyrannical oratory, Cicero divides his own speeches in the same four categories as those of his Greek model (i.e.,

88 There is only one extant passage in which Cicero compares himself explicitly with Demosthenes: Cic. Orat. 105, cf. n. 92 below. Cicero’s extreme adoration for the Greek orator is a feature of his later work, appearing for the first time in 46 BC: see e.g. Brut. 35, Orat. 6, cf. sections 1.4 (on the relative insignificance of Demosthenes in Cic. De or. ), 2.2 (on Demosthenes’ as an icon of Atticism) and 2.3.2 (on Cicero’s praise of Demosthenes’ style). Obviously, the Roman author modeled the fourteen speeches of his Phil. (44–43 BC) on Demosthenes: see Wooten (1977), Dugan (2005) 309–314, and Usher (2008), who argues that Dem. De cor. was the most important source of inspiration for Cic. Phil. Bishop (2016) discusses the parallels that Cicero draws between his own career and Demosthenes’ life in his Brut. and Orat. : see n. 90 below. It was a common feature of ancient literary criticism to compare the life, the oratory and the style of Cicero and Demosthenes, and to appoint a winner: already in the first century BC, Caecilius of Caleacte composed such a ‘comparison’ (σύγκρισις ), cf. T6 Woerther (= IX Ofenloch). In the first and second centuries AD, such comparisons can be found in Long. Subl. 12.4, Quint. Inst. orat. 10.1.106–109, Gell. NA 15.28.6–7, and Plut. Dem. 3. According to Dugan (2005) 315– 332, the comparisons testify to ‘Cicero’s success in his challenge to Demosthenes’ supremacy in the history of eloquence’. De Jonge (2018b) notes that the debate on the relative superiority of the two orators ‘reenacts as it were the imaginary competition between Demosthenes and Cicero’. 89 Evidence for Demosthenes’ significance as a political icon in Rome comes mainly from imperial declamations: see Bishop (2016) 172–173. The declamations on Demosthenes are collected by Kohl (1915) 66– 82; most of these texts focus on Demosthenes’ opposition to Philip after the Athenian defeat at Chaeronea (338 BC). Note that both Prop. 3.21.27 and Petr. 5 refer to the orator’s words as the ‘weapons of Demosthenes’ (Demosthenis arma ), indicating that his verbal defense of Greek freedom was considered an important part of his legacy. On Cicero’s view that he is one of the last free orators before the dark age of tyranny, see Brut. 22, 330. 90 Bishop (2016) 183. Cicero gives an autobiographical description of his career in Brut. 313–324. 91 Relevant passages are, e.g., De cor. 60–72, 95–101, 172, 199–208, 219–221, 235–236, 245–247, 277–278.

221

CHAPTER FIVE grand, simple, intermediate and varied), adding quasi-modestly: ‘You are certainly aware that where I try many things, he brings many things to perfection, that where I have the desire to speak in whatever way a case demands, he has the ability.’ 92 Admittedly, the imitation of Demosthenes does not leave much room for optimism about the fate of the Roman republic: after all, even the most celebrated Attic orator was ultimately unable to prevent Philip’s victory and Athens’ downfall as a free city-state. Yet, in 46 BC, Cicero still believed that he could save the republic, not so much as an active orator, but rather as a teacher and as a canonical author. He emphatically exhorts his young friend Brutus to pick up the baton by studying his various orations alongside the speeches of Demosthenes: by presenting his own oeuvre as mandatory reading material for a new generation, Cicero aims to secure the survival of his passionate oratory after his imminent exit from the stage of Roman oratory and politics. 93 Through the benefit of hindsight, we now know that Brutus did not heed Cicero’s instructions: he eventually chose real daggers over the proverbial weapons of Ciceronian eloquence. Cicero himself, however, never stopped believing in the power of his own exuberant oratory: after the assassination of Caesar, he briefly returned to the center stage of history as the author of fourteen vehement Philippics against the triumvir Marc Antony (44–43 BC).94 Although they led to his vicious decapitation and to the exhibition of his severed head and hands on the Rostra, these speeches testify once again to Cicero’s unwavering conviction that grand oratory, modeled after the elevated eloquence of Classical Athens, could save Rome from tyranny.

92 Cic. Orat. 105: Vides profecto illum multa perficere, nos multa conari, illum posse, nos velle quocumque modo causa postulet dicere. In his categorizations of his own speeches and those by Demosthenes, Cicero seems to make his own personal canon correspond almost point by point to his Greek predecessor’s: in Orat. 102 and 111, he presents his Caec. as the equivalent of Dem. Lept. in the simple style, he mentions his own Rab. Post. as the counterpart of Dem. Phil. in the grand style, and he aligns his Cluent. and Corn. to Dem. De cor. and Fals. leg. among the varied speeches. Cicero gives his own Leg. Man. as an example of a speech composed in the intermediate style; he does not offer us examples of such speeches from the Demosthenic corpus. 93 Dugan (2005) 251–332 and Bishop (2016) 185–190 argue that Cicero uses the figure of Demosthenes to secure his own oratorical legacy: according to Bishop, ‘Cicero the literary critic has given the Roman reading public a guide to reading and interpreting Cicero the orator’. I will only add that Cicero is not just concerned about his place in the literary canon, but that he also still trying to influence Roman politics by advertizing his passionate rhetoric to a new audience, including (not in the last place) young Brutus. In Brut. 22, Cicero dramatically hands over his scepter to Brutus. Cic. Orat. 105 presents Brutus as an exemplary student of both Demosthenes and Cicero: ‘You have him (i.e., Demosthenes) constantly in your hands, and yet also find time to read my speeches over and over’ ( nec eum dimittis e manibus et tamen nostra etiam lectitas ). 94 On Cic. Phil. , see the literature cited in section 2.2 n. 19 above.

222

TO BE ATTIC OR NOT TO BE ATTIC

5.5.2 Dionysius’ Middle Style and the Rule of Augustus Notwithstanding Cicero’s confidence in the thunderlike force of grand oratory, the republican constitution of Rome inevitably crumbled under the weight of civil war. Some years later, during the early days of the imperial era, Dionysius articulates his views about prose style: as we already saw, he favors a mixed type of style that holds an appropriate mean between two opposite extremes (section 3.5). According to him, such a well-balanced mean not only constitutes the culmination of brilliant literature, but it also stands for ‘excellence in life and in conduct’ ( ἀρετὴ τῶν βίων καὶ τῶν ἔργων ).95 In the present section, I will focus on the moral and political dimensions of Dionysius’ Atticism: we will see that his conception of Attic style in general and the middle style in particular resonates (1) with his views about civic virtue in Classical Greece as well as early and contemporary Rome, and (2) with the political program of the emperor Augustus. Dionysius’ approach to prose style, then, is not only inseparably connected to his ethics and politics, but it also resonates with the behavior of Rome’s supreme ruler. The close connections between Dionysius’ stylistic views on the one hand and his ideas about proper civic behavior on the other hand are palpable in his entire extant oeuvre. We have already seen that the preface to On the Ancient Orators presents Attic oratory under the guise of an exemplary citizen—a faithful wife (σώφρων γαμετή ), who judiciously takes care of her household (section 5.4). In his critical works, Dionysius discusses various passages that praise the σωφροσύνη of the Athenians (usually referring to their ‘self- restraint’),96 while other citations from his favorite Attic orators evoke moral principles like ‘piety’ ( εὐσέβεια ), ‘justice’ ( δικαιοσύνη ) and manly ‘valor’ ( ἀνδρεία ).97 It seems that

95 Dion. Hal. Comp. 24.2. Dionysius cites ‘Aristotle and the other philosophers of his school’ ( Ἀριστοτέλης τε καὶ οἱ ἄλλοι ὅσοι κατ ’ ἐκείνην τὴν αἵρεσιν φιλοσοφοῦσιν ) as the authorities for the view that excellence (ἀρετή ) is a mean in all aspects of human activity. See section 3.5 n. 110–118 for the Peripatetic background of Dionysius’ stylistic theories. 96 On the flexible meaning of σωφροσύνη and its cognates, see section 5.3 n. 50 and section 5.4 n. 73 above: in the case of Dion. Hal. Ant. orat. 1.5, I translated σώφρων as ‘faithful’, whereas I chose ‘sound’ to render the same word in Ant. orat. 1.7. For the present section, which discusses σωφροσύνη as an attribute of virtuous men, I will assume the word to refer to temperance and self-restraint: cf. Rademaker (2007) 257–260. This interpretation can be readily connected to Dionysius’ conception of the middle style as well-balanced and properly measured. Moreover, the Greek scholar often replaces σωφροσύνη in political contexts with the related term ἐγκράτεια (‘self-control’): see e.g. Ant. Rom. 2.10.4, 2.28.2. 97 On these four virtues in Dionysius’ historiographical and critical works, see Goudriaan (1989) 207–210, who argues that ‘the traces of political-ethical theory in Dionysius’ work point in the same direction as the

223

CHAPTER FIVE

Dionysius purposely selects passages that advertise these virtues; in On Isocrates , Dionysius even expatiates at length on the civic virtues that the eponymous orator evokes in his middle- style prose, where ‘the best possible lessons in virtue can be found’ ( κράτιστα δὴ παιδεύματα πρὸς ἀρετὴν ἔστιν εὑρεῖν). According to the Greek critic, Isocrates’ Panegyricus , for instance, instills in its audience a sense of ‘civic excellence’ ( πολιτικὴ καλοκἀγαθία ), reminding them that the old Greeks were not only ‘formidable warriors’ ( τὰ πολέμια δεινοί ), but also ‘noble- hearted’ (τὰ ἤθη γενναῖοι ), ‘covetous of honor’ (φιλότιμοι ) and ‘self-controlled’ (σώφρονες ), always keeping an eye to ‘moderation’ ( μετριότης ); Dionysius moreover posits that the Letter to Philip spurs not just the Macedonian king but all rulers to ‘excellence’ ( ἀρετή ), that On the Peace is a cogent exhortation ‘to justice and piety’ ( ἐπὶ τὴν δικαιοσύνην καὶ τὴν εὐσέβειαν ), and that Areopagiticus forces any reader to become a ‘more orderly citizen’ (κοσμιώτερος ).98 These virtues, as we have seen, are typically associated with Classical Athens (section 5.3). Dionysius’ stylistic views closely correspond to his views about civic excellence: after all, his most beloved type of style is characterized by attributes such as ‘balance’ ( συμμετρία ), ‘timing’ ( εὐκαιρία ), ‘due measure’ ( μέτρον ) and ‘appropriateness’ ( πρέπον ), befitting the pious, thoughtful and restrained behavior of his ideal citizen, personified in the faithful

philosophical elements of his rhetorical and stylistic theory’ (‘de sporen van politiek-ethische theorie in Dionysius’ werk wijzen in dezelfde richting als de wijsgerige elementen in zijn rhetorische en stijlkritische theorie’). See also Wiater (2011) 67: ‘The result was a world view in which the Greeks represented a set of moral and political virtues, the most important of which were ἐλευθερία , δικαιοσύνη , εὐσέβεια , ἀνδρεία , and σωφροσύνη .’ On the relationship between Greece and Rome in Dionysius’ work, see not only section 1.5 above, but also the present section n. 108–109 below. In addition to the passage from Isoc. discussed in the next note below, Dionysius quotes various passages from Attic authors that explicitly refer to the four virtues mentioned above: for σωφροσύνη , see Isoc. 17.1, Dem. 3.3, 43.3, Comp. 23.19, 25.19; for εὐσέβεια , see Dem. 1.2, 21.2, Thuc. 33.2, Pomp. 6.6; for δικαιοσύνη , see Isoc. 17.1, Dem. 30.2, Pomp. 6.6; for ἀνδρεία , see. Dem. 26.8, 30.2, Thuc. 12.2, Amm. II 17.2. Cf. n. 100 below for the attestations of these terms in Ant. Rom. Concerning Dionysius’ selection of sample passage in his critical works, see also section 2.2 on Dion. Hal. Comp. 18.26 (= Hegesias F5 Jacoby): he quotes the passage from Hegesias’ work not only on stylistic grounds but also on the basis of its content, as it depicts the deplorable behavior of a Babylonian (or, ‘Asian’) king. 98 Dion. Hal. Isoc. 4.3, 5.1–3, 6.3, 7.1, 8.1. These statements all belong to Dionysius’ summary of the four aforementioned Isocratean speeches in Isoc. 4–8. According to Wiater (2011) 71–77, the passage presents ‘a list of what Dionysius regards as the key elements of Classical identity’, giving Isoc. ‘the character of a “handbook of Classical identity”, which provides the readers with a standardized, easily accessible definition of what it means to be Classical. Thus, Dionysius’ essay demonstrates that Classical rhetoric is not simply a certain rhetorical style among others, but is coupled with a well-defined set of political and moral values.’ On Wiater’s view that Dionysius aims to create a separation between Greek and Roman readers, see n. 109 below.

224

TO BE ATTIC OR NOT TO BE ATTIC housewife who acts as his Attic muse. 99 Strikingly, the qualities of Dionysius’ mixed style not only correspond to the moral and political values that he finds embodied in Classical Greek literature: in his Roman Antiquities , too, he praises the early Romans for engaging in similarly virtuous conduct. In the second book of his history, for example, Dionysius dwells on the constitution of king Romulus, underlining several virtues that will be familiar to the students of Dionysius’ critical essays, viz., ‘self-restraint’ ( σωφροσύνη ), ‘piety’ ( εὐσέβεια ), ‘justice’ (δικαιοσύνη ), ‘valor’ ( ἀνδρεία ) and various other qualities.100 Moreover, Dionysius assigns an important place to oratory in the virtuous society of primeval Rome: ‘By persuading and informing one another, by yielding in some things and gaining other things from their opponents, who yielded in turn, they settled their disputes in a manner befitting fellow citizens.’ 101 To be brief, oratorical excellence and civic virtue go hand in hand in Dionysius’ understanding of both Classical Greece and archaic Rome. Likewise, deterioration of oratory is closely connected to political and moral decline: in Dionysius’ view, this equally holds good for post-Classical Greece and for post-regal Rome. 102

99 See esp. Dion. Hal. Dem. 34.5 with the discussion of Dionysius’ mixed style in section 3.5 above. 100 For Romulus’ constitution, see Dion. Hal. Ant. Rom. 2.1–29. See, specifically, ibid. 2.18.2: Romulus ‘recognized that good laws and the emulation of worthy pursuits render a state pious, temperate, devoted to justice, and brave in war’ ( ἔγνω διότι νόμοι σπουδαῖοι καὶ καλῶν ζῆλος ἐπιτηδευμάτων εὐσεβῆ καὶ σώφρονα καὶ τὰ δίκαια ἀσκοῦσαν καὶ τὰ πολέμια ἀγαθὴν ἐξεργάζονται πόλιν ). Cf. the discussion in Goudriaan (1989) 360– 381 and Wiater (2011) 172–185, who emphasize the Greekness of Romulus’ constitution. Several scholars, such as Pohlenz (1924), Gabba (1960) and Ferrara (1970), have proposed that Dionysius’ description of Romulus’ constitution actually derives from a political pamphlet (‘Tendenzschrift’) composed in the time of Sulla (Gabba), Caesar (Pohlenz) or Augustus (Ferrara). Yet, we should note that Dionysius’ description of the virtues in Romulus’ laws neatly ties in with his admiration for Classical Attic prose style and culture: therefore, it seems more expedient to interpret the passage in the light of Dionysius’ own overarching project. Cf. the objections to the pamphlet theory in Balsdon (1971). Delcourt (2005) offers a clear overview of Dionysius’ passage on Romulus’ constitution and the relevant secondary literature. The four virtues mentioned above are often referred to throughout Ant. Rom. : for σωφροσύνη , see 1.5.3, 2.74.1, 2.75.1, 6.59.1; for εὐσέβεια , see ibid. 1.4.2, 2.62.5, 3.17.2–3, 11.43.3; for δικαιοσύνη , see ibid. 4.9.9, 8.61.2–3, 9.53.6, 10.57.3; for ἀνδρεία , see ibid. 4.3.1, 5.25.4, 6.6.1, 9.17.4. Cf. n. 97 above for several attestations of the same virtues in Dionysius’ criticial treatises. 101 Dion. Hal. Ant. Rom. 2.11.3: Ἀλλὰ πείθοντες καὶ διδάσκοντες ἀλλήλους καὶ τὰ μὲν εἴκοντες , τὰ δὲ παρ ’ εἰκόντων λαμβάνοντες , πολιτικὰς ἐποιοῦντο τὰς τῶν ἐγκλημάτων διαλύσεις . 102 For Dionysius’ idealization of the moral heydays under Romulus, see Fox (1996) 53–63 and (2019) 193–196; for the historian’s views about the subsequent decline under later kings and especially during the early Republic, see Pelling (2019) 215–218. According to Dion. Hal. Ant. Rom. 2.11.3, the Romans continued to use oratory as a means to settle disputes until the tribuneship of C. Gracchus (123–122 BC), from which point onward they ‘have been perpetually slaying and banishing one another from the city and refraining from no irreparable acts in order

225

CHAPTER FIVE

The consistent emphasis on the same basic virtues throughout Dionysius’ historiographical and critical works raises an important question: what message does the Greek author aim to get across about the politics of Augustus? It has proven impossible to reach a consensus on Dionysius’ political allegiance on the basis of his historiographical work: the Greek scholar mentions the emperor only once by name (merely as a means to date his own arrival in Rome), and it is extremely tricky to compare passages from his Roman Antiquities (which do not seem to convey a clear-cut message about monarchical government) to the young emperor’s political program (which seems to have been in flux in the early years of his principate). 103 Still, it is undeniable that the themes that the Greek scholar addresses, like exempla of virtuous behavior, regime change and the relationship between Greece and Rome, are highly topical in Augustan Rome: we can safely assume, therefore, that Dionysius taps into a commonly available discourse on politics and society, in the same way as his critical treatises are built on a shared discourse on stylistic theory.104 For a more straightforward picture of Dionysius’ approach to Augustan politics, however, we should set aside the ambiguous allusions of his historiographical work and instead turn to the preface of On the Ancient Orators.

to gain the upper hand’ ( οὐκέτι πέπαυνται σφάττοντες ἀλλήλους καὶ φυγάδας ἐλαύνοντες ἐκ τῆς πόλεως καὶ οὐδενὸς τῶν ἀνηκέστων ἀπεχόμενοι παρὰ τὸ νικᾶν). Thus, Dionysius lets Rome’s long era of grim revolutions and civil wars coincide with oratorical decline. 103 Dion. Hal. Ant. Rom. 1.7.2 mentions Augustus’ victory in the civil war: cf. section 1.1 n. 2 above. Pro- Augustan readings of Ant. Rom. focus on Dionysius’ positive statement about the gens Iulia (1.70.4), and on the similarities between Romulus’ constitution and Augustus’ policy: see esp. Martin (1971), Luraghi (2003) 275, and Delcourt (2005) 297. Anti-Augustan interpretations have been proposed by Hill (1961) 90–91, Hurst (1982) and Gabba (1982a) 801, who argue that Dionysius’ stress on the Greek origins of Rome contradicts the focus on the city’s Italic origins in the Latin literature of the Augustan era. Pelling (2019) 219–220 is more reserved, noting that ‘many of the “Augustan” resonances can be taken either way, with hints of reservation or criticism as well as of acclaim’; cf. Pelling’s discussion of Ant. Rom. 10.50–60 on the decemvirate of 451–449 BC. According to Wiater (2011) 206–216, ‘Dionysius’ image of the early Romans is not at odds with Augustan ideology in particular, but with contemporary Roman conceptions of Roman identity and attempts of upper-class Romans to distinguish themselves from the Greeks’. In my view, however, Dionysius makes a connection (rather than a separation) between Greeks and Romans: see esp. n. 108–109 below. On the difficulties of testing Augustan-era texts against Augustus’ imperial propaganda, see esp. D.F. Kennedy (1992). 104 Cf. Delcourt (2005) 369: ‘L’ombre du principat plane sur les Antiquités Romaines .’ See also Pelling (2019) 219–220: ‘There are so many places where we might catch a contemporary whiff. (…) We should think of Dionysius echoing those preoccupations of Augustan political discourse without preaching about them, introducing ideas that meshed with the political propaganda without crudely echoing it or taking sides.’

226

TO BE ATTIC OR NOT TO BE ATTIC

The introductory essay is known as a ‘manifesto of classicism’: 105 it tells the story of a renaissance, not only celebrating the revival of Attic oratory but also the rebirth of the concomitant political and moral values. Here, Dionysius is unequivocally positive about the Rome of his day: he expresses his gratitude to the ‘age in which we live’ (ὁ καθ ’ ἡμᾶς χρόνος ) and specifically to ‘the conquest of the world by Rome’ (ἡ πάντων κρατοῦσα Ῥώμη ) for bringing about a revolutionary change for the better. I have quoted the relevant passages below. 106

Πολλὴν χάριν ἦν εἰδέναι τῷ καθ ’ ἡμᾶς χρόνῳ δίκαιον , ὦ κράτιστε Ἀμμαῖε, καὶ ἄλλων μέν τινων ἐπιτηδευμάτων ἕνεκα νῦν κάλλιον ἀσκουμένων ἢ πρότερον , οὐχ ἥκιστα δὲ τῆς περὶ τοὺς πολιτικοὺς λόγους ἐπιμελείας οὐ μικρὰν ἐπίδοσιν πεποιημένης ἐπὶ τὰ κρείττω . (...) Αἰτία δ’ οἶμαι καὶ ἀρχὴ τῆς τοιαύτης μεταβολῆς ἐγένετο ἡ πάντων κρατοῦσα Ῥώμη πρὸς ἑαυτὴν ἀναγκάζουσα τὰς ὅλας πόλεις ἀποβλέπειν καὶ ταύτης δὲ αὐτῆς οἱ δυναστεύοντες κατ ’ ἀρετὴν καὶ ἀπὸ τοῦ κρατίστου τὰ κοινὰ διοικοῦντες , εὐπαίδευτοι πάνυ καὶ γενναῖοι τὰς κρίσεις γενόμενοι , ὑφ’ ὧν κοσμούμενον τό τε φρόνιμον τῆς πόλεως μέρος ἔτι μᾶλλον ἐπιδέδωκεν καὶ τὸ ἀνόητον ἠνάγκασται νοῦν ἔχειν .

We ought to acknowledge a great debt of gratitude to the age in which we live, my most accomplished Ammaeus, for an improvement in certain fields of serious study, and especially for the considerable revival in the practice of civil oratory. (…) I think that the cause and origin of this great revolution has been the conquest of the world by Rome, who has thus made every city focus its entire attention upon her. Her leaders are chosen on merit ( κατ ’ ἀρετήν ), and administer the state according to the highest principles (ἀπὸ τοῦ κρατίστου ). They are thoroughly cultured (εὐπαίδευτοι ) and noble in their judgment (γενναῖοι τὰς κρίσεις ), so that under their ordering influence (ὑφ’ ὧν κοσμούμενον ) the sensible section (τὸ φρόνιμον μέρος ) of the population has increased its power and the foolish have been compelled to behave rationally (νοῦν ἔχειν ).

105 The designation is coined by Hidber (1996): ‘das klassizistische Manifest’. Cf. section 1.5 n. 82 above. Hidber (esp. ibid. 75–81) argues that Dionysius was a pro-Augustan proponent of cultural integration between Greeks and Romans, with Rome as the world’s educational capital. See also Spawforth (2012) 20–26 on the political dimensions of Dionysius’ account of Atticism and Asianism. 106 Dion. Hal. Ant. orat. 1.1, 3.1. The passage is partly quoted and discussed in section 1.6 above.

227

CHAPTER FIVE

Dionysius endows the present ‘leaders’ ( δυναστεύοντες ) of Rome with the very virtues that lie at the core of the Classical Attic prose that they revived. Both the ancient Athenians and the modern Romans, for instance, are described as ‘noble’ ( γενναῖοι ), and they are both preoccupied with ‘excellence’ ( ἀρετή ). Also, while the Attic muse is described as ‘philosophical’ ( φιλόσοφος ), the Roman leaders are praised as ‘sensible’ ( φρόνιμοι ); and while Isocrates is said to make his readers ‘more orderly citizens’ ( κοσμιώτεροι ), Rome’s leaders are presented as ‘having an ordering influence’ ( κοσμούμενον ).107 Hence, Dionysius’ interpretation of the middle style with its emphasis on temperance and balance is not only well-equipped as a vehicle for the civic discourses of the Classical era, but it is also perfectly at home in Augustan Rome. All in all, Dionysius’ coherent praise for Athenian culture, Attic prose, early Roman politics and contemporary Roman culture seems to forge a strong unity between Greeks and Romans: he presents both groups as having contributed to a shared golden age of outstanding prose style and civic excellence—the Greeks as original creators, the Romans as conquerors and reinventors.108 I cannot, at any rate, accept the view, proposed by Nicolas Wiater, that Dionysius is engaged in a polemic with his Roman contemporaries, rubbing their noses in the superiority of Greek culture. 109

107 Cf. especially the quotations from Dion. Hal. Isoc. 4–8 in n. 98 above, and the passage on the Attic muse (Ant. orat. 1.5–7) quoted in section 5.4 above. 108 For the notion of a renaissance shared by Greeks and Romans alike, cf. Dion. Hal. Ant. orat. 3.2: many fine oratorical, historiographical and philosophical works ‘have proceeded as the products of well-directed zeal from the pens of both Greeks and Romans, and will probably continue to do so’ ( καὶ Ῥωμαίοις καὶ Ἕλλησιν εὖ μάλα διεσπουδασμέναι προεληλύθασί τε καὶ προελεύσονται κατὰ τὸ εἶκος ). We should not assume, as Usher (1974) 10 and Wisse (1995) 77 seem to do, that Dionysius refers specifically to Calvus and his Atticist movement as the driving force behind the revival of Attic oratory: after all, we have seen in section 5.2 above that authors with diverse stylistic preferences fashioned themselves ‘Attic’ without associating themselves with each other. In the absence of clear proof, Grube (1965) 212 divines that Dionysius refers to Cicero and Caesar, while Heldmann (1982) 125 argues that he is thinking of the Latin literature from the Ciceronian and Augustan periods. In addition to the city’s cultural achievements, Dionysius also addresses Rome’s ‘violent imperialism’, as Whitmarsh (1998) notes: ‘The reason he gives for the revival of oratory's fortunes is 'Rome's conquest (κρατοῦσα ) of everything, her coercion (ἀναγκάζουσα ) of all cities to look at her, and the fact that her despots (δυναστεύοντες ) manage public affairs virtuously and most excellently (ἀπὸ τοῦ κρατίστου , which picks up the earlier use of the κράτος root).’ The forceful language does not necessarily criticize Rome: in Ant. Rom. 1.3.5, Dionysius accepts Rome’s violent dominion of the world as legitimate. 109 Following such studies as Gabba (1982b), Wiater (2011) argues that the foremost concern of Dionysius’ historiography and criticism is the assertion of Greek superiority over Rome: cf. section 1.2 n. 26 and the present section n. 103 above. Wiater ibid. 100 connects the reference to Rome’s world dominion to Alexander’s fight

228

TO BE ATTIC OR NOT TO BE ATTIC

Still, the question remains whom exactly Dionysius is praising, when he commends Rome’s ‘leaders’ ( δυναστεύοντες ) for being chosen on merit and for governing the state according to the highest principles: does Dionysius have Augustus and his ministers in mind, or rather a larger group of influential aristocrats, such as the Tuberones and the Metilii, with whom he associated himself after his arrival in the city?110 I consider it almost inconceivable that Dionysius would not include the emperor among Rome’s praiseworthy leaders: indeed, the (self-styled) reputation of Augustus appeals to the very moral and stylistic ideals that Dionysius advertises. The emperor’s policies and propaganda are focused on Augustus’ role in restoring peace, justice, religion and morality; Dionysius’ outrage at the Asian ‘concubine’ (ἑταίρα ) expelling the Attic ‘wife’ ( γαμετή ) tallies with Augustus’ austere laws on adultery and celibacy. 111 Furthermore, Suetonius emphasizes that the emperor was ‘extremely temperate’ (continentissimus ) in his personal life, shunning excess: his house and furniture are ‘modest’ ( modicus ), while his diet was ‘paltry’ (minimus ) and ‘plain’ ( vulgaris ). 112 In other against the Persians: ‘Dionysius thus re-interprets the crucial event in contemporary Roman history, Augustus’ principate, as a turning point in the prolonged, Greek struggle against the Barbarian.’ While Wiater rightly dwells on Dionysius’ view that the Romans are indebted to Greek culture, we cannot but conclude that Ant. orat. contains nothing to proclaim Greece’s superiority; if anything, he offers all the credits for the revival of Attic oratory to Rome, and he refers to both Greek and Roman authors, without imposing any intrinsic hierarchy (cf. the previous note). Cf. Spawforth (2012), who awards a large role to Rome in the revival of Classical Greek culture, aptly referring to the ‘re-hellenization’ of Greece from the West. 110 Most scholars think that the word δυναστεύοντες refers to Augustus or (as it is a plural) to Augustus and his ministers: see e.g. Wilamowitz (1900) 45, Bonner (1939) 10, Kennedy (1972) 352, De Jonge and Hunter (2019b). Goudriaan (1989) 574 thinks that Dionysius refers to ‘the city of Rome as a whole, specifically the Roman elite’ (‘de stad Rome als geheel, de Romeinse elite in het bijzonder’); Hidber (1996) 119–120 suggests that Dionysius has the Roman aristocrates in mind who acted as patrons for Greek and Roman authors. For Dionysius’ connections to the Tuberones and Metilii, see section 1.5 above. 111 There is no shortage of modern studies on Augustus and his program of restoration. Zanker (1988) 101–166, for instance, reviews Augustus’ ‘program of cultural renewal’ by studying Augustan visual arts and architecture. Cf. more recently, e.g., Eder (2005), who focuses on the balance that Augustus sought between raising his profile and renouncing personal power, and Richardson (2012), who similarly explores the emperor’s ‘restoration of the Republic’ versus his ‘establishment of the Empire’. On the connection between Dionysius’ Atticism and Augustus’ marital laws, see De Jonge (2014a) 397. Cf. Schulze (2019) 174–176, who argues that the depiction of women and families in Dion. Hal. Ant. Rom. resonates with Augustus’ legislation, specifically his Lex Iulia de Maritandis Ordinibus (18 BC) and Lex Iulia de Adulteriis Coercendis (17 BC). According to Galinsky (1981), the laws underline Augustus’ concern for the assertion of Rome’s domination over other peoples. 112 Suet. Aug . 61–78 discusses the emperor’s ‘private and domestic life’ ( interior ac familiaris eius vita ). The four adjectives quoted here are taken specifically from Aug . 72.1, 73, 76.1 and 77 respectively.

229

CHAPTER FIVE cases, his conduct is described as holding a mean between two extremes: ‘he was no less strict than gracious and merciful’ toward his slaves and his clients; ‘his togas were neither close nor full, his purple stripe neither narrow nor broad’; and he celebrated holidays ‘lavishly as a rule, but sometimes only in a spirit of fun’. 113 Suetonius also tells us that Augustus initially lived near the Forum in a small house which had belonged to the orator C. Licinius Macer Calvus, and that he later moved to the Palatine hill into the equally modest dwelling of the orator Q. Hortensius Hortalus.114 It is hard not to read this move as a metaphor for the emperor’s moderate, well-balanced character, and ditto literary taste. After all, as I will discuss in greater detail below (section 5.6.1), Calvus advocated an extremely sober oratorical style, which he advertised as quintessentially ‘Attic’, whereas Hortensius exhibited a more lavish rhetoric, which was commonly regarded as ‘Asian’. Thus, Augustus could be associated both with a leading Atticist and a prominent Asianist from the recent past: the princeps himself naturally occupies a middle ground between these polar opposites. Similarly, Suetonius’ description of the emperor’s ‘oratorical style’ ( genus eloquendi ) is not a far cry from Dionysius’ mixed stylistic register: Augustus’ speech is described as ‘elegant and moderate’ ( elegans et temperatum ), holding a judicious mean between Asian-style ‘wannabe innovators’ ( cacozeli ) such as his friend Maecenas, and bland ‘archaizers’ ( antiquarii ) such as his stepson Tiberius. 115 Augustus is said to carefully

113 Suet. Aug. 67.1: Non minus severus quam facilis et clemens. Ibid. 73: Togis neque restrictis neque fusis, clavo nec lato nec angusto. Ibid. 75: Profusissime, nonnumquam tantum ioculariter. The toga could be used as a metaphor for oratorical style: see Sen. Ep. 114.4 (with section 5.6.1 n. 139 below) on the wantonness of loose togas, and Tac. Dial. 26.1 (with section 4.5.1 n. 89 above) on the solemnity of rugged togas. 114 Suet. Aug. 72.1. Wardle (2014) 453–454 identifies the orator Hortensius as the former occupant of the Palatine house; Suetonius carefully avoids associating Augustus with Hortensius’ extravagant oratory by claiming its modesty ( aedes modicae ). De Jonge (2018a) 169–171 observes that Suetonius’ description of Augustus’ Palatine dwelling appeals to a taste for ‘noble simplicity’ that was widely shared in Augustan Rome: a similar preference is exemplified in Dion. Hal. Comp. 3.7–12 (on Eumaeus’ shed in Hom. Od. 16.1–16) and Verg. Aen. 8.364–369 (on king Evander’s modest hut on the Palatine). 115 Suet. Aug. 86.1–3. The translations are mine. Giordano (2000) 36–45 notes that the term temperatum designates a balance between the extremes of Atticism and Asianism, connecting the term elegans to the stylistic virtues of correctness and clarity, exemplified in the prose of Julius Caesar. Cf. the division of elegantia in Rhet. Her. 4.17 into ‘clarity’ (explanatio ) and ‘correct Latin’ ( Latinitas ); see also section 5.4 n. 64 above. Suetonius, however, seems to stress the emperor’s obsession with clarity at the expense of correctness: it was Augustus’ chief aim ‘to express his thought as clearly as possible’ ( sensum animi quam apertissime exprimere ), to which end he used prepositions with the names of cities, while he also repeated conjunctions. For a careful discussion of the emperor’s elegantia on the basis of the extant fragments and testimonies, see Gutiérrez Gonzalez (2012).

230

TO BE ATTIC OR NOT TO BE ATTIC avoid the ‘meaningless phrases and artificial arrangement’ (sententiarum ineptiae atque concinnitas ) and the ‘perfume-dripping locks’ ( myrobrechis cincinni ) of the former group, whilst equally steering clear of the ‘stink of far-fetched words’ ( reconditorum verborum fetores ) of the latter group. He reserves his biggest scorn for Marc Antony, whom he considers ‘insane’ ( insanus ) for not being able to find a proper mean between the faulty opposites, hopping from one extreme to the other, resulting in a style that is unsuitable for serious subject matter. 116 The Roman emperor Augustus and the Greek immigrant Dionysius, then, share a fondness of moderation and propriety, both in speech and in action. This section has focused on the connections between Dionysius’ middle style on the one hand and his preoccupation with such civic virtues as self-restraint, piety, justice and valor on the other hand. We have seen that his favorite type of style is perfectly equipped to express the outstanding political and moral values of Classical Athens, which he also recognizes as the defining features of both Romulean and Augustan Rome. According to Dionysius, Greeks and Romans alike share in the current glorious renaissance of oratory, politics and morality. The correspondences between Dionysius’ project and Augustus’ self- presentation, lastly, show that the ethics and aesthetics of the middle met with great support in the highest echelons of Roman society: in a remarkable way, the Greek scholar epitomizes the zeitgeist of Augustan Rome, as the emperor himself liked to see it.

5.6 Calvus’ Atticism: Challenging Rome’s Rhetorical and Political Elites Finally, we will now turn back the clock to the middle of the first century BC, when mighty Augustus was still a child named C. Octavius. In oratory, Cicero and Hortensius were the most celebrated speakers of the Forum; in politics, the alliance of Caesar, Pompey and Crassus held sway over the city. It is in this period that C. Licinius Macer Calvus makes a

De Jonge (2019) 264 connects Suetonius’ description of Augustus’ style to Dion. Hal. Ant. orat. 3.1: ‘When Dionysius expresses his gratitude to Rome by observing that its leaders are εὐπαίδευτοι (well educated) and γενναῖοι τὰς κρίσεις (excellent in their judgment), it is therefore tempting to conclude that Dionysius is (also) thinking of the eloquence of Augustus himself.’ Augustus’ style (intermediate between archaizing and baroque) roughly matches Dionysius’ middle style (intermediate between simple and elevated, or between rough and smooth). Cf. sections 3.5 and 4.5.3 on archaisms and modernisms in Dionysius’ threefold divisions of style. 116 Augustus’ criticism of Antony’s ‘perverse and inconsistent taste’ ( malum et inconstans iudicium ) echoes the fierce scorn that Hegesias receives for not being able to create a style that is suitable for serious discourse: in accordance with Agatharchides’, Cicero’s and Dionysius’ criticism of Hegesias, Antony is censured by Augustus for composing ‘pieces of writing rather to be admired than to be understood’ ( ea quae mirentur potius homines quam intellegant ). Cf. section 2.4, esp. 2.4.1 n. 114 above.

231

CHAPTER FIVE

‘meteor-like’ appearance on the firmament of Latin literature. 117 Like a meteor, he only lit up briefly: the record of his activities is limited to the period between 56 and 54 BC. 118 Very few fragments of his oratory and poetry are extant. 119 He died not later than 47 BC, still in his twenties or early thirties (section 1.4). Yet, much like a meteor, Calvus left an awesome impression: as a neoteric poet, he was considered on a par with his famous friend Catullus; as an orator, he was one of the most formidable challengers of Cicero. Calvus is the only author whose name can be linked indubitably to the so-called Atticist movement that was active in the late 50s and early 40s BC (section 5.2). In this section, I will focus on the interplay of politics, ethics and stylistic theory in the extant record of Calvus, for which I will turn not only to the sources for his activities as an orator and a politician, but also to the few surviving fragments of his poetry, and to the testimonies on his private life. Despite the scanty evidence, we will clearly see that Calvus’ views on Attic style are primarily designed for polemical purposes: focusing on the typically Attic values of purity and masculinity (section 5.3), he challenges at once the rhetorical and political elites of his day. First, we will see that Calvus’ conception of Atticism is tailored to attack the exuberant eloquence of Rome’s most successful orators Cicero and Hortensius (section 5.6.1). Subsequently, we will see that his doctrines do not only challenge Rome’s predominant oratorical style, but also the behavior of the city’s most powerful politicians, specifically Caesar and Pompey (section 5.6.2). Thus, Calvus presents Athens as an antidote against the decadence and degeneracy that, in his view, dominated the courts and the assemblies on the Forum. In addition, we will see that Calvus upheld his Attic ideal both in public and in his private life—possibly even during his sleep.

5.6.1 Calvus against the Rhetorical Status Quo Often regarded—by ancient and modern scholars alike—as the prototype of Roman Atticism (section 5.2), the significance of Calvus as a key player in the history of Roman rhetoric is

117 Münzer (1926) 428: ‘C. Licinius Macer Calvus ist eine Erscheinung, die am Himmel der römischen Literatur meteorgleich aufleuchtete und einen starken allgemeinen Eindruck, aber im einzelnen nur wenige deutliche Spuren hinterliess.’ 118 The brief window onto Calvus’ activities opens with his prosecution of Asitius in 56 BC (Cic. Cael. 23, Tac. Dial. 21), and it closes with his head-to-head clash with Cicero at the trial of Vatinius in 54 BC (Sen. Controv. 7.4.6, Quint. Inst. orat. 9.2.25, Tac. Dial. 21). For the known details about Calvus’ life, see Castorina (1946). 119 Malcovati (1967) 492–500 collects the fragments of Calvus’ oratory; Hollis (2007) 49–86 presents and discusses his twenty-odd extant poetical snippets. He was famous for composing an epyllion under the title Io.

232

TO BE ATTIC OR NOT TO BE ATTIC obvious. Yet, the student of Calvus’ oratory should be aware of two important caveats that complicate our interpretation of his views and achievements. First of all, ancient sources often conflate Calvus with his Atticist movement: they ascribe to the man the same views as to the entire movement. Therefore, we may cautiously use the extant notes that speak in general about Calvus’ Atticist group as sources for his own views.120 Unfortunately, next, our knowledge of Calvus’ stylistic program is built to a large extent on the work of his mortal enemy Cicero: Calvus, in the words of the elder Seneca, ‘for a long time waged a most hostile contest with Cicero for the supremacy in oratory’. 121 Unsurprisingly, Cicero presents a rather cynical account of the motivations and goals underlying Calvus’ stylistic program. Thus, we should be aware of the inherent bias in the treatment of Calvus and his Atticist allies by Cicero and by later authors that (partly) build their judgment on Cicero’s works. Still, it is worthwhile to make ourselves familiar with the prejudice in Cicero’s discussion, as it testifies to the polemic context of Calvus’ stylistic views. 122 All in all, the famous rhetorician offers roughly four explanations for Calvus’ meager style in his Brutus and Orator . First, he suggests that Calvus’ motives were elitist: his dry oratory was only appreciated by a small group of connoisseurs, but it did not appeal to the palate of the masses. According to Cicero, this led Calvus and his Athenophile friends to the misguided conclusion that they were in fact ‘wise men’ ( prudentes, intellegentes ). 123 Secondly, Cicero submits that the Atticists subjected themselves to anxious self-scrutiny, avoiding at all cost to make mistakes: this over-scrupulousness rendered their style, in Cicero’s words, ‘bloodless’ (exsanguis ) and ‘sapless’ ( exsucus ). 124 Thirdly, Cicero compares Calvus’ movement to a cult: the self-styled Atticists considered simplicity and lack of ornament to be ‘a venerable and

120 The conflation of Calvus and his movement is apparent in Cic. Brut. 284: Calvus ‘was in error and caused others to err with him’ ( et ipse errabat et alios etiam errare cogebat ). Moreover, the extant sources name no other orators who belong to Calvus’ movement of Atticizing speakers: see section 5.2 above. 121 Sen. Controv. 7.4.6: Diu cum Cicerone iniquissimam litem de principatu eloquentiae habuit. 122 Dugan (2001) 412–413 offers an overview of the terms that Cicero and Calvus use in their polemic, showing that they deploy dichotomies focusing, e.g., on swelling and emaciation, extravagance and restraint, sickness and health. On the personal relationship between Calvus and Cicero, see the literature cited in n. 128 below. 123 Cic. Orat. 236, for instance, claims that his Atticizing opponents considered themselves ‘wise’ ( prudentes ) on account of their refusal to adorn their discourse with well-knit rhythms; Cic. Brut. 289, however, notes that the Atticists’ speeches were so boring that no-one even bothered to attend. On the close association between intellectualism and the simple style in the Roman rhetorical tradition, cf. section 3.6 above. 124 Cic. Brut. 283. Cf. Orat. 76 on the lack of vigor of Calvus’ Atticism, and Rhet. Her. 4.16, which designates the faulty style akin to the simple register as ‘bloodless’ ( exsanguis ). See also section 3.4 n. 81–82 above.

233

CHAPTER FIVE almost religious obligation for the orator’ ( quasi religio et verecundia oratoris ). 125 Lastly, Cicero mockingly notes that the Atticists ‘measure the power of eloquence by their own ability’. In other words, they adhere to the simple style not by choice, but by necessity: they are simply not skilled enough to attain grandeur and magnificence. 126 To summarize, Cicero submits that snobbishness, extreme caution, religious zeal and incompetence are the pillars on which Calvus’ conception of Atticism is built. Cicero’s diatribe should, of course, not be taken at face value. Later discussions of Calvus’ style in the works of the elder Seneca, Tacitus and Quintilian are inevitably affected by Ciceronian bias, but they add several perspectives that might complement or even correct Cicero’s account. I will come back to the testimony of Seneca at the end of the present section; Tacitus and Quintilian, to begin with, attest to the popularity of Calvus in the late first and early second centuries AD. Quintilian, for one, reports that even his in day he found men ‘who preferred Calvus above all others’. In addition, Tacitus still had access to Calvus’ twenty-one books of ‘little speeches’ ( oratiunculae ): he reports that the speeches against Vatinius even served as textbooks for students of rhetoric. 127 Apparently, Quintilian and Tacitus were both familiar with Calvus’ (now lost) correspondence with Cicero. 128 This

125 Cic. Brut. 284. Cf. also Quint. Inst. orat. 12.10.14, who compares Calvus’ Atticist movement to a ‘clique of initiates in some sort of mystery’ ( manus quasi quibusdam sacris initiata ), who treat Cicero as an ‘unbeliever’ (parum superstitiosus ). 126 Cic. Opt. gen. 10: Vim eloquentiae sua facultate metiuntur. Cf. Orat. 234, where Cicero explains that orators who did not adopt a rhythmical style, did so, because ‘they could not attain this’ ( hoc assequi non potuerunt ). Tac. Dial. 21.2 makes a similar statement: Aper claims that Calvus’ lack of elegance and elevation should not be attributed to any autonomous ‘choice’ (voluntas ), but rather to a lack of ‘intellectual force’ ( ingenium ac vires ). 127 Quint. Inst. orat. 10.1.115: Inveni qui Calvum praeferrent omnibus. Tac. Dial. 21.1–2: ‘Calvus himself, although he left behind as many, if I am right, as twenty-one volumes, hardly comes up to the standard in any one, or two at the most, of his speeches. (...) Yet, the orations entitled Against Vatinius are a common textbook with students’ ( ipse mihi Calvus, cum unum et viginti, ut puto, libros reliquerit, vix in una et altera oratiuncula satis facit ... At hercule in omnium studiosorum manibus versantur accusationes quae in Vatinium inscribuntur ). Calvus’ speeches against Vatinius (54 BC) attracted the attention of later readers, because (1) they featured a direct confrontation with Cicero, cf. Gruen (1967); (2) they pertained to Caesar, who had bought Vatinius’ services; and (3) Calvus’ performance in the trial was the topic of a joke in Catull. 53. Cf. section 5.6.2 below on Catullus’ poem and on Calvus’ political views. See Gruen (1971) on the circumstances of the trial. For the popularity of Calvus in the early second century AD, see also Plin. Ep. 1.2.2, who admires the figures of speech of ‘my beloved Calvus’ ( meus Calvus ). 128 The mutual letters are referred to in Quint. Inst. orat. 9.4.1, 12.1.22, and Tac. Dial. 18. Hendrickson (1926) thinks that the letters were primarily concerned with Cicero’s and Calvus’ opposing views on word arrangement.

234

TO BE ATTIC OR NOT TO BE ATTIC material shows us that Calvus indeed deliberately cultivated a meager style, but for different reasons than the ones that Cicero suggests: Calvus and his supporters devised a stylistic program that was intended to be diametrically opposed to the abundant eloquence that, to their minds, unduly dominated Rome. Tacitus presents not only Calvus’ views of Ciceronian oratory, but also the opinions of Brutus, who was a close friend of Cicero’s, but who may also have been sympathetic to Calvus’ restrained oratory.129

Satis constat ne Ciceroni quidem obtrectatores defuisse, quibus inflatus et tumens nec satis pressus, sed supra modum exsultans et superfluens et parum Atticus videretur. Legistis utique et Calvi et Bruti ad Ciceronem missas epistulas, ex quibus facile est deprehendere Calvum quidem Ciceroni visum exsanguem et aridum, Brutum autem otiosum atque diiunctum; rursusque Ciceronem a Calvo quidem male audisse tamquam solutum et enervem, a Bruto autem, ut ipsius verbis utar, tamquam ‘fractum atque elumbem.’

Even Cicero himself, as is well known, had his detractors: they thought him turgid and puffy, wanting in conciseness, inordinately exuberant, redundant, and not Attic enough. You have read, of course, the letters of Calvus and Brutus to Cicero, from which it is easy to gather that, as for Calvus, Cicero thought him bloodless and attenuated, just as he thought Brutus spiritless and disjointed; while Cicero was in turn criticized by Calvus as relaxed and unmanly, and by Brutus, to use his own words, as ‘broken and emasculated.’

Gruen (1967) remarks that, despite their obvious differences of opinion in political and oratorical matters, the two men underheld a cordial relationship based on mutual respect. Although Cicero and Calvus may indeed have exchanged occasional pleasantries, Dugan (2001) 426, however, rightly emphasizes the fierceness of their enmity: ‘At stake in Calvus’ and Cicero’s polemics was their fundamental selfhood, which accounts for the fervor of their contestation.’ In section 5.6.2 below, we will see that Calvus’ Atticism not merely pertained to his views on oratory, but that it affected every aspect of his life. 129 Tac. Dial. 18.4–5. Cic. Att. 15.1a.2 gives a clue concerning Brutus’ sympathies for Calvus’ Atticism, when Cicero emphatically claims that Demosthenes’ thunderbolts are thoroughly Attic, a point which he also makes in Orat. 23, 234: the passage is quoted in section 2.3.2 n. 60 above. Shackleton Bailey (1999) 374 argues that Cicero’s correspondence with Brutus incited the former to compose his Brutus. Dugan (2005) 261, likewise, claims that Brutus’ Atticist tendencies made him an appropriate recipient of both Cicero’s Brutus and his Orator. Yet, there is no obvious connection between Brutus and Calvus, except for the fact that they both criticized Cicero. Even if Brutus appealed to Attic style, there is no good reason to subscribe him Calvus’ entourage: cf. the literature mentioned in section 5.2 n. 15 above.

235

CHAPTER FIVE

Quintilian was also familiar with the letters that Cicero exchanged with his opponents about oratorical style. In his discussion of the contents of this correspondence, he does not show himself as impartial as Tacitus: he repeats, for instance, Cicero’s view that Calvus’ movement was a pseudo-religious cult. Yet, even without mentioning Calvus by name, Quintilian offers us some reliable information about his views, consistent with the passage in Tacitus’ Dialogus .130

At M. Tullium non illum habemus Euphranorem circa pluris artium species praestantem, sed in omnibus quae in quoque laudantur eminentissimum. Quem tamen et suorum homines temporum incessere audebant ut tumidiorem et Asianum et redundantem et in repetitionibus nimium et in salibus aliquando frigidum et in compositione fractum, exultantem ac paene, quod procul absit, viro molliorem (...). Praecipue vero presserunt eum qui videri Atticorum imitatores concupierant. Haec manus quasi quibusdam sacris initiata ut alienigenam et parum superstitiosum devinctumque illis legibus insequebatur.

But in Cicero we have one who is not just an Euphranor, distinguished in several branches of art, but a man supreme in everything for which anyone wins praise. 131 And yet his own contemporaries had the hardihood to attack him as too bombastic, Asian, redundant, too repetitive, sometimes frigid in his humor, and broken, exuberant and —which should be unthinkable—almost softer than a man in his arrangement. (...) It was those who wanted to be thought imitators of the Attic writers who were particularly hard on him. Like a clique of initiates in some sort of mystery, this group attacked him as an outsider and unbeliever, and refusing to be bound by their rules.

The testimonies of Tacitus and Quintilian present a clear picture of the objections of Calvus and others against Cicero’s style. The gist of their criticism is that Cicero was ‘not Attic enough’ ( parum Atticus ), but rather a ‘foreigner’ ( alienigena ) and even ‘Asian’ ( Asianus ). 132

130 Quint. Inst. orat. 12.10.12–14. On Cicero’s correspondence with Calvus en Brutus, see Inst. orat. 9.4.1 and ibid. 12.1.22 131 Euphranor (fl. 364–361 BC) was a painter and sculptor from Corinth, who also wrote theoretical works on symmetry and color. Plin. NH 35.129 tells us that he was famous for the versatility of his talent. 132 According to Delarue (1982), the stylistic epithet ‘Asian’ was invented by Calvus specifically for his polemic against Cicero. While it is difficult to prove this thesis, it is true that, as a term of abuse, the adjective ‘Asian’ seems to have become relevant for the first time in the controversy between Cicero and the self-styled Attic

236

TO BE ATTIC OR NOT TO BE ATTIC

The vocabulary and imagery that is used to articulate this view merit a closer look: the anti- Ciceronian slurs in the passages above can be divided into roughly three groups. First of all, Tacitus and Quintilian both report that the self-proclaimed Atticists take offense at Cicero’s unrestrained verbosity: he was ‘not concise enough’ (neque satis pressus ), ‘redundant’ (redundans ), ‘repetitive’ (in repetitionibus nimius ), ‘inordinately exuberant and overabundant’ (super modum exsultans et superfluens ). 133 Briefly put, Cicero’s style is thought to be characterized by excess, which, as we have seen, is traditionally recognized as a hallmark of Asia (section 5.3). In ancient literary criticism, likewise, futile verbosity is often associated with Asianism: Cicero himself, for instance, asserts that Asian oratory often contains ‘certain meaningless words inserted to fill out the rhythm’ (inania quaedam verba quasi complementa numerorum ). Brevity, conversely, is usually presented as a defining feature of the simple style, which Calvus cultivated as the only truly Attic style. 134 Secondly, Cicero’s style is presented as a disease: it is ‘inflated’ ( inflatus ) and ‘swollen’ ( tumens, tumidior ). The medical references are connected to the analogy that many rhetoricians and critics observed between rhetorical style and the human body. 135 In this case, Cicero is diagnosed with a malignant tumor, evoking an image that is often used to criticize authors who ineptly adopt a grand, ornate style: the fourth book of Rhetorica ad Herennium ,

orators surrounding Calvus: see section 5.2 above. Hence, Lucarini (2015) may be right in claiming that Cicero himself invented the twofold division of Asian oratory ( Brut. 325) as a reaction to the sudden prominence of Asianism in the rhetorical discourse. See section 1.6 above and n. 143 below for Cicero’s take on Asian style. 133 Verbosity also comes up in the comparisons ( συγκρίσεις ) between Cicero and Demosthenes that start to appear briefly after Cicero’s death: see section 5.5.1 n. 88 above. Quint. Inst. orat. 10.1.106, for instance, notes that ‘nothing can be subtracted’ ( nihil detrahi potest ) from Demosthenes’ speeches, while to Cicero’s orations ‘nothing can be added’ ( nihil addi potest ). Long. Subl. 12.4, in addition, compares Cicero’s style to a ‘widespread conflagration’ ( ἀμφιλαφής τις ἐμπρησμός ), while Demosthenes is said to resemble a thunderbolt. 134 Cic. Orat. 230. On the redundancy of Asian style, see also Suet. Aug. 86.3, who quotes Augustus mocking the ‘verbose and unmeaning fluency of the Asiatic orators’ ( Asiaticorum oratorum inanis sententiis verborum volubilitas ). On the connection between brevity and the simple style, see e.g. Dion. Hal. Lys. 4.4–5 and section 3.2.1, table 4 above. 135 On the pervasiveness of anatomical imagery in stylistic discussions, see section 3.2 n. 21 above and the literature cited there. We already saw (e.g., Quint. Inst. orat. 12.10.58) that the terminology of the three styles is to a large extent based on the metaphor of the human body: the grand style can be described as ‘stout’ ( ἁδρός , robustus ), while the simple style can be referred to as ‘slim’ ( ἰσχνός , subtilis ). Dugan (2001) 413 underlines the importance of corporeal metaphors in the polemic between Cicero and Calvus: ‘The body in rhetoric cannot be marked off as purely metaphorical or purely literal: rhetoric is a system of thought that constantly returns to the issue of the bodily self of the orator.’ Cf. section 5.6.2 below for Calvus’ treatment of his own body.

237

CHAPTER FIVE for instance, includes a passage on the ‘inflated style’ ( sufflata figura ), which is presented as the vicious counterpart of the grand register. The author draws an explicit comparison between this faulty style and a sick body: ‘Just as a swelling ( tumor ) often resembles a healthy condition of the body, so, to those who are inexperienced, turgid and inflated language ( ea quae turget et inflata est ) often seems majestic.’ 136 Physical health, then, plays a large role in the debate between Calvus and Cicero: both men describe their own style in terms such as ‘health’ ( sanitas ), ‘wholesomeness’ ( salubritas ) and ‘soundness’ ( integritas ), while they present their opponent’s style as mortally ill. Specifically, Calvus’ sparse oratory could be written off as emaciated, while Cicero’s abundantly decorated speeches could be regarded as bloated. 137 The third prominent theme in the anti-Ciceronian discourse of Calvus and his followers is effeminacy, another widely acknowledged attribute of Asian culture and oratory (section 5.3). Cicero received abuse for being ‘broken’ (fractus ), ‘loose’ (solutus ), ‘sinewless’ (enervis ) and ‘loinless’ (elumbis )—words that each carry strong feminine connotations. The words ‘sinew’ (nervus ) and ‘loin’ (lumbus ), for instance, are slangy expressions that can denote male genitalia: thus, the letters of Brutus and Calvus quite literally depict Cicero as emasculated. 138 The words ‘loose’ (solutus ) and ‘broken’ (fractus ), additionally, are also connected to effeminacy and degeneracy. 139 The orators in Rome frequently impugned each

136 Rhet. Her. 4.15: Ita ut corporis bonam habitudinem tumor imitatur saepe, item gravis oratio saepe inperitis videtur ea quae turget et inflata est. On bloating and inflation as a fault associated with grand oratory, see also Hor. Ars. P. 27, who reports that ‘an author who strives for grandeur, is swollen’ ( professus grandia turget ); Long. Subl. 3.4 on ‘tumor’ ( ὄγκος ); and Gell. NA 6.14.5 on ‘inflated and swollen speakers’ ( sufflati atque tumidi ). Cf. also Gutzwiller (1969) and section 3.2.1 n. 26 above. 137 For Cicero’s and Calvus’ self-diagnosed good health, see esp. Cic. Brut. 51, 284, with the synoptic table in Dugan (2001) 412–413. According to Dugan, Calvus and Cicero have at their disposal a ‘system of tropes based on the body and its regulation’ to defend their own speech and to attack each other’s oratory. 138 Hendrickson (1926) 255–258 reckons the terms elumbis and enervis —‘an offense to the Roman ear’—among the most hostile epithets in the polemic between Cicero and his opponents. For an overview of sexual and gendered vocabulary in Roman rhetoric, see Adams (1982) 149–151, who explains, for instance, that nervus can mean both ‘tendon’ and ‘penis’, since ‘the penis could be regarded as a tendon or group of tendons’. 139 For a discussion of these words as technical rhetorical terms, see Richlin (1997) 94. Sen. Ep. 114.4 applies the term ‘loose’ ( solutus ) to the effeminate style of Maecenas: ‘Is his style not as loose as the man himself is unguirded?’ ( non oratio eius aeque soluta est quam ipse discinctus ). Cf. Corbeill (1996) 160 n. 81 on the connection between unguirdedness and femininity. The word ‘broken’ ( fractus ), next, also typically appears in conjunction with words denoting emasculation: see, e.g., Sen. Suas. 2.23 on ‘extreme ornamentation and broken arrangement’ ( nimius cultus et fracta compositio ); Sen. Vit. beat. 13.4, referring to the pleasure-seeking type of

238

TO BE ATTIC OR NOT TO BE ATTIC other’s masculinity, usually provoked by their opponent’s high-pitched voice, theatrical performance or careful for word arrangement. The jibes against Cicero’s manhood focus on the latter aspect of his oratory: Quintilian, as we have seen, indignantly reports that Cicero was thought ‘broken’ (fractus ) and ‘almost softer than a man’ (paene viro mollior ) in the arrangement of his words. Cicero himself was aware of this critique: he notes that there are orators who think that ‘speech is emasculated ( enervatur ) by a careful arrangement of words’. 140 The Roman Forum was, of course, a men’s world: attacking Cicero’s masculinity could be an effective strategy to defame his position as Rome’s leading orator. 141 Thus, to Calvus’ mind, Ciceronian eloquence symbolized three crucial stylistic errors —unrestrained verbosity, unhealthy bombast and unmanly word arrangement. When Calvus enters our record in 56 BC, Cicero was still the principal orator of Rome, but he was more vulnerable on account of his recent exile (58–57 BC) and his diminished political position in the face of Caesar and Pompey: he was, therefore, an obvious target for Calvus’ contentious spirit. Yet, Cicero’s grand oratory was part of a wider trend in Roman eloquence: Q. Hortensius Hortalus (115–50 BC), whose reputation as a passionate orator was second only to Cicero’s, was susceptible to the same criticisms as the ones that Calvus and his friends

man as ‘weakly, broken, losing his manhood, and on the sure path to baseness’ ( enervis, fractus, degenerans viro, perventurus in turpia ); and Quint. Inst. orat. 1.10.31 on ‘effeminate’ ( effeminata ) music, ‘broken down by indecent rhythms’ ( impudicis modis fracta ). Note that the word ‘broken’ often refers to word arrangement; as we saw in section 2.3.2 above, Cic. Brut. 386 censures Hegesias’ composition for being ‘broken’ ( fractus ) and ‘chopped-up’ (minutus ), while Orat. 170 accuses the self-styled Atticists of composing sentences that are ‘broken down and minced’ ( infracta et amputata ); in Orat. 235, by contrast, Cicero advertises his own rhythm as ‘well-knit’ ( aptus ). Thus, Cicero could readily defend himself against charges of effeminacy. 140 Cic. Orat. 229. Cf. ibid. 231: an orator ‘will weaken his rhythms’ ( delumbet sententias ), if he hunts for short, monotonous jingles. Cf. also Sen. Ep. 115.2: ‘Verbal balance is not a masculine ornament’ ( non est ornamentum virile concinnitas ). On the gendered approach to oratorical performance ( actio ), see esp. Richlin (1997) 99–105. Examples are not hard to come by, e.g., in Cic. Orat. 54: ‘There should be no effeminate bending of the neck’ (nulla mollitia cervicum ) and the orator should check himself ‘through a manly attitude of his body’ ( virili laterum flexione ). Rhet. Her. 3.22 claims that a sharp exclamation irritates the audience, as it is more similar to a ‘womanish shriek’ ( muliebris vociferatio ) than to ‘manly dignity’ ( virilis dignitas ). 141 Gleason (1995) 71–75, 98–102 points out that the Forum was a logical setting for invectives focusing on virility, as it was the place of every Roman boy’s transition to manhood. Richlin (1997) 106–107 focuses especially on the role of gender in the rhetorical schools, arguing that gendered language flowed from these schools into the controversies between adult orators. Connolly (2010) 88, lastly, approaches the role of gender in the rhetorical discourse from a social-political perspective: ‘What is at stake is the inculcation and perpetuation of a particular set of attitudes and behaviors associated with masculinity and men.’

239

CHAPTER FIVE launched against Cicero. 142 In his Brutus , Cicero sets Hortensius up as the paradigm of ‘Asian style’ (Asiatica dictio ), presenting him as a master of the two subtypes of Asianism—the ‘pithy and clever kind’ ( genus sententiosum et argutum ) and the ‘swift and impetuous kind’ (genus volucre atque incitatum ). 143 Although Cicero’s discussion is quite benevolent to Asian oratory, he says much about Hortensius’ style that could give offense to Calvus: according to Cicero, Hortensius ‘made a study of gracefully pointed phrases’, that were ‘merely graceful and of pleasant sound, though not necessary nor always useful’, his prose lacked weight, wherefore it was better suited to youth than to old age. 144 It is obvious, then, that Hortensius does not meet Calvus’ standards of purity and restraint: if anything, he may even be further removed from Calvus’ Attic ideal than Cicero. We also know that Hortensius was mocked for his alleged femininity. He was known to take great care in arranging his toga and he tended to gesture excessively during his performances: according to some, he behaved more like an actor than like an orator, more like a woman than like a man. 145 Aulus Gellius tells us the following anecdote about his

142 Cf. section 5.5.2 above on Hortensius’ Palatine dwelling, which would later become the house of Augustus. 143 Cic. Brut. 325. Cicero’s treatment of Asian style is somewhat ambivalent: he criticizes it for lacking authority (auctoritatis parum ), being ‘more condoned in youth than in old age’ (adulescentiae magis concessum quam senectuti ), but in his description of the two subtypes of Asianism he adopts a rather kind approach. This positive attitude may have several reasons: (1) as we saw in section 1.6 above, Cicero does not want to disavow the lessons of his Asian teachers, esp. Menippus of Stratonicea, Dionysius of Magnesia, Aeschylus of Cnidus and Xenocles of Adramyttium ( Brut. 315–316); (2) by focusing on the youthful spirit of Asianism, Cicero can explain why Hortensius was a successful orator as a young man, though he has fallen from grace later in life; (3) as Vasaly (2002) 86 points out, Cicero can account for his own fondness for playful stylistic ornaments in his early speeches; and (4) Cicero could lecture his opponents, who use the epithet ‘Asian’ as a term of abuse, on the complexities of Asian oratory. Lucarini (2015) carefully examines Cicero’s discussion of the two Asian styles and the styles of the authors that he connects to them (for the first type Timaeus, Hierocles and Menecles; for the second type Aeschylus of Cnidus and Aeschines of Miletus); given the polemic context of Cicero’s discussion, however, we should be aware that Cicero’s descriptions might not match any real stylistic practices. 144 Cic. Brut. 326–327: Habebat (…) illud studium crebrarum venustarumque sententiarum, in quibus (…) erant quaedam magis venustae dulcesque sententiae quam aut necessariae aut interdum utiles. According to Dyck (2008), Cicero grew to respect his formal rival Hortensius for introducing to the Roman Forum a passionate style and for taking oratory seriously as a career. Yet, Garcea and Lomanto (2014) point out the surreptitious reservations that Cicero always keeps with respect to Hortensius, already present in the prediction of his success in De or. 3.228–230, which was modeled on the announcement of Isocrates’ success in Pl. Phaedr. 229a–230b. 145 See Gell. NA 1.5.2: ‘Because he dressed with extreme foppishness, Hortensius arranged the folds of his toga with great care and exactness, and in speaking used his hands to excess in lively gestures, was assailed with jibes and shameful charges; and many taunts were hurled at him, even while he was pleading in court, for appearing

240

TO BE ATTIC OR NOT TO BE ATTIC performance during the trial of Sulla, the dictator’s nephew (62 BC), when his opponent Lucius Torquatus called him ‘Dionysia’, after a notorious dancing girl. Hortensius reportedly replied to this taunt by saying ‘in a soft and gentle voice’ (voce molli atque demissa ): ‘I would rather be a Dionysia, Torquatus, indeed a Dionysia, than like you un étranger aux Muses, à Aphrodite, à Dionysos’ (ἄμουσος , ἀναφρόδιτος , ἀπροσδιόνυσος ). 146 Hortensius, then, does not deny the charges of effeminacy and preciosity, but he proudly embraces them: he repeats his new nickname twice, he answers his opponent in Greek and he even adopts a woman’s voice. He almost mentions the girl’s name a third time in the final word ἀπροσδιόνυσος , a clever pun meaning both ‘unconnected to Dionysus’ and ‘not to the point, mal à propos’. 147 This quaint anecdote shows us that Hortensius’ playful speeches are, even more than Cicero’s oratory, prone to Calvus’ Atticist criticisms. Thus, the movement led by Calvus was first and foremost a movement of protest, largely directed against the dominant oratorical style in Rome. Under the banner of Attic style, Calvus proposed to replace verbosity with brevity, bombast with sobriety, and effeminacy with masculinity—in a word, he sought to drive out the degeneracy of Asia in favor of the purity and restraint of Athens. Yet, it seems that not even Calvus himself always lived up to his stylistic promises: the elder Seneca reports that he was a ‘violent and passionate performer’ (violentus actor et concitatus ) and that he arranged his words after the example of Demosthenes. The epilogue to his speech for Messius even displays, in Seneca’s view, a ‘soft’ (emollita ) and ‘womanly’ (infracta ) composition. 148 According to Tacitus, like an actor’ ( quod multa munditia et circumspecte compositeque indutus et amictus esset manusque eius inter agendum forent argutae admodum et gestuosae, maledictis compellationibusque probris iactatus est, multaque in eum, quasi in histrionem, in ipsis causis atque iudiciis dicta sunt ). Cic. Orat. 132 and Quint. Inst. orat. 11.3.8–9 posit that Hortensius’ writings are better than his performances; according to Val. Max. 8.10.2, the actors Roscius and Aesopus studied Hortensius’ gestures. Cf. Steel (2007) 243–244, who regrets our dependence on Cicero’s testimonies on Hortensius’ oratory, while ‘the details resist discovery’. 146 Gell. NA 1.5.3: Dionysia, Dionysia malo equidem esse quam quod tu, Torquate , ἄμουσος , ἀναφρόδιτος , ἀπροσδιόνυσος . For French translations to render Greek words in Latin passages, cf. section 4.2. n. 22 above on Lucil. fr. 74–75. Gunderson (2000) 127–131 and Williams (2010) 172–174 discuss Hortensius as an example of an orator who willfully bends the rules of masculinity in Rome. The three Greek words at the end of the sentences are used by Hortensius to attack Torquatus’ unrefined character (esp. his unfamiliarity with the pleasures of ‘Wein, Weib und Gesang’); Gellius agrees with this charge, calling Torquatus a ‘man of somewhat boorish and uncouth nature’ ( subagresti homo ingenio et infestivo ). 147 For the proverbial meaning of ἀπροσδιόνυσος , see, e.g., Cic. Att . 16.13a.1. 148 Sen. Controv. 7.4.6–8. To illustrate Calvus’ passionate performance, Seneca reports that the orator, in the spur of the moment, used to rise up from his benches and run to his opponent’s side of the court. In Calvus’

241

CHAPTER FIVE likewise, the second part of his speech against Vatinius was ‘ornate in words and sentences’ (verbis ornata et sententiis ) and ‘adapted to please the ears of the judges’ (auribus iudicum adcommodata ). 149 These testimonies are blatantly at odds with Calvus’ reputation as a meager and dry orator. Yet, the inconsistency teaches us an important lesson about Calvus’ Atticism: it was principally devised to challenge the dominant orators of Rome and usurp their leading position. Calvus’ oratory was obviously more restrained than the style of his opponents, but his meagerness was by no means as extreme as his polemical discourse suggests.

5.6.2 Calvus against Public and Private Effeminacy Calvus’ program of Atticism is not only designed to undercut the authority of the leading orators of his day: his fervent campaigns for restraint, purity and masculinity are also expressions of a penetrating social critique aimed to provoke the city’s political elite. 150 In brief, Calvus was troubled by the indecent, effeminate behavior of Rome’s ruling class: he therefore not merely sought to change the dominant oratorical style of the Forum, but he also wanted to correct the underlying moral degeneracy, specifically targeting Caesar and Pompey, but certainly not sparing himself. This subsection, then, will address the relationship between Calvus’ conception of Attic style and his professed desire, expressed particularly in the extant portions of his poetry, to install an austere moral code in Rome. Thus, we will see that rhetoric, poetry, politics and ethics are inextricably connected in Calvus’ Atticism. As far as we know, Calvus never held any political office nor do his oratorical activities reveal his allegiance to any politician. Yet, his activities as a forensic lawyer inevitably involved him in Roman politics: in the year 54 BC alone, for instance, he defended C. Messius, whom Pompey considered a friend, he prosecuted P. Vatinius, one of Caesar’s protégés, and he spoke for C. Porcius Cato, a triumviral partisan. 151 In addition to his oratorical fame, Calvus was also an accomplished poet, a prominent sidekick to Catullus, who

Demosthenic composition, moverover, Seneca claims that ‘there is nothing that is sedate or gentle, everything is excited and stormy’ ( nihil in illa placidum, nihil lene est, omnia excitata et fluctuantia ). On the connection between Calvus’ Atticism and Demosthenes, see also Plin. Ep. 1.2.2 and section 5.2 n. 22 above. 149 Tac. Dial. 18.2. 150 Cf. Delarue (1982) 178–179, who argues that such designations as ‘broken and loinless’ ( elumbis ), that we saw in Tac. Dial. 18.4–5 (section 5.6.1 above), do not only pertain to oratorical style, but also to moral character. 151 Gruen (1967) expertly discusses Calvus’ involvements in the trials of Messius, Vatinius and Cato, paying close attention to the political implications: ‘Calvus was born to politics.’ Calvus’ father was a people’s tribune, praetor, historian and orator: cf. Sall. Hist. 3.48. Gruen suggests that Calvus’ defense of Cato is the fruit of his reconciliation with Caesar after they had fallen out over Calvus’ anti-Caesarian verses: see n. 157 below.

242

TO BE ATTIC OR NOT TO BE ATTIC jokingly recounts how one of his listeners calls Calvus an ‘eloquent dwarf’ ( disertum salaputium ), punning on his small physical stature and perhaps also on his restrained style. 152 There are two noteworthy connections between Calvus’ Atticizing oratory and his neoteric poetry. First, the slenderness of his ‘little verses’ ( versiculi ) matches the meagerness of his ‘little speeches’ ( oratiunculae ): in both literary pursuits, he shunned wordiness, pompous grandeur and effeminacy. 153 Secondly, in his poetry, he is outspoken about the faults of contemporary politicians: in two much-cited fragments, he mocks Caesar and Pompey on account of their luxuriousness and effeminacy. 154 The moral flaws that Calvus identifies in the city’s leading senators are strikingly similar to the stylistic vices for which he censures Cicero. I quote Calvus’ two extant poetical bashings of the triumvirate below, starting with his ‘notorious verses’ ( notissimi versi ) on Caesar. 155

... Bithynia quicquid et pedicator Caesaris umquam habuit.

... all that Bithynia and Caesar’s bugger ever possessed.

152 Catull. 53: ‘A fellow in the crowd made me laugh just now: when my dear Calvus had drawn out in splendid style his accusations against Vatinius, he lifted up his hands in wonder, saying ‘Great gods, what an eloquent dwarf!’’ ( Risi nescio quem modo e corona, / qui, cum mirifice Vatiniana / meus crimina Calvus explicasset, / admirans ait haec manusque tollens, / Di magni, salaputium disertum! ). Keith (1999) 45 suggests that Catullus’ reference ‘may participate in contemporary Atticist polemic’. On Calvus’ diminutive posture, see Sen. Controv. 7.4.7: ‘He was a small man’ ( erat parvolus statura ). Catullus contrasts the petiteness of Calvus’ physique to the awesome effects of his oratory: cf. Fordyce (1961) 223–225. Sen. Controv. 7.4.6 relates that Tigellius, who was prosecuted by Calvus, exclaimed during trial: ‘I ask you, judges: just because he is eloquent, must I be convicted?’ ( rogo vos, iudice, num, si iste disertus est, ideo me damnari oportet? ). Deroux (2008) focuses on the hapax salaputium , translating it as ‘tiny dick’ (‘petite queue’). See Catull. 14, 50, 96 for other references to Calvus’ oratory and poetry. 153 Keith (1999) 42–46 discusses the similarities between Catullus’ and Cicero’s literary critical vocabulary. Male sexuality plays a central role in the verses of both Calvus and Catullus: cf. Wray (2001). For the dislike of lengthy and grand poetry in Alexandrian poetry and in Catullus’ circles, see Johnson (2007). There is also an important difference between Calvus’ oratorical style and his poetics: he used Attic prose of the Classical era as his model for his speeches, while he admired the Alexandrian poets of the Hellenistic era as a neoteric poet, cf. Bowersock (1979) 62–63. 154 To the fragments quoted below, add fr. 36 Hollis (= fr. 3 Courtney), which mocks the venality of a certain Tigellius, a known friend of Caesar. Cf. Hollis (2007) 80: ‘Calvus’ attacks on him may have been meant to discredit Caesar too.’ 155 Calvus fr. 38 Hollis (= fr. 17 Courtney = Suet. Iul. 49.1).

243

CHAPTER FIVE

This partly surviving elegiac couplet refers to Caesar’s stay in Bithynia at the court of king Nicomedes IV (81–80 BC): Caesar dallied so long in Asia that it was rumored that he had a homosexual affair with the king. The story offered his friends and opponents ample ammunition to insult him: he was called ‘Queen of Bithynia’ (Bithynica regina ) and his soldiers famously compared his conquest of Gaul to Nicomedes’ conquest of Caesar. 156 The original context of Calvus’ verses is unknown: they could be part of a discussion about a man who enriched himself from the fall of the Bithynian kingdom, or they could simply be a hyperbolic description of a vast sum of money. 157 It is not hard to see why the story was of interest to the Attic-minded Calvus: Caesar had not only stayed in Asia, but he could also be associated with a supposedly sinful Asian life-style. Calvus’ words effectively embarrassed Caesar: Calvus even felt obliged to seek reconciliation. 158 The most damaging part of the poem seems to be the reference to Caesar’s role as a passive sexual partner: Calvus rudely calls Nicomedes ‘Caesar’s bugger’ (pedicator Caesaris ), signaling the king’s dominance over a weak and effeminate Caesar. In a similar vein, Calvus lashed out at Pompey:159

Magnus, quem metuunt omnes, digito caput uno scalpit; quid dicas hunc sibi velle? virum.

Magnus, whom everyone fears, scratches his head with one finger. What would you say he is after? A man!

The epigram, which may be the only of Calvus’s poems to survive completely, mocks Pompey’s dainty habit of running a single finger through his hair, a gesture which was held to

156 Suet. Iul. 49 presents the fragment from Calvus in a list of taunts that refer to the scandalous affair with Nicomedes, culminating in the song of his own soldiers on the occasion of his triumph over Gaul (46 BC). 157 Hollis (2007) 81–83: ‘The identity of the profiteer has been forgotten, but everyone remembered the insult to Julius Caesar, which may even have been aggravated by the fact that it was merely a passing reference.’ 158 Suet. Iul. 73: ‘When Gaius Calvus, after some famous lampoons, took steps through his friends toward a reconciliation, Caesar wrote to him first and of his own free will’ ( Gaio Calvo post famosa epigrammata de reconciliatione per amicos agenti ultro ac prior scripsit ). The fact that Suetonius refers to Calvus’ ‘famous lampoons’ ( famosa epigrammata ) suggests that he may have composed a series of poems with jibes against Caesar. On Calvus’ quarrel and reconciliation with Caesar, see esp. Gruen (1967) 224–225. Suetonius (ibid.) notes that Catullus, too, attacked and subsequently made amends with Caesar. 159 Calvus fr. 39 Hollis (= fr. 18 Courtney = schol. Iuv. 9.133, schol. Luc. 7.726). Juvenal’s scholiast refers to the couplet as a complete epigram ( tale epigramma ). Both scholiasts attribute it to Martial, which cannot be correct: Sen. Controv. 7.4.7 already quotes part of the extant words (from digito onwards), ascribing it to Calvus.

244

TO BE ATTIC OR NOT TO BE ATTIC be a sign of effeminacy and homosexuality. 160 Plutarch also discusses Pompey’s little tick, denying that it indicated anything untoward: ‘An unwarranted suspicion of unmanliness (μαλακία ) was aroused against Pompey on account of his habit of scratching his head with one finger, although he was far removed from effeminacy ( θηλύτης ) and licentiousness (ἀκολασία ).’ 161 As in Caesar’s case, the stories about Pompey’s effeminate behavior may have been prompted by his prolonged stay in Asia, where he spent the better part of the 60s BC. Calvus, in any case, has little doubt about Pompey’s intentions: he wants a ‘man’ (vir )! The latter word refers specifically to a ‘dominant sexual partner’, putting Pompey, like Caesar, in the passive role. 162 Calvus, then, was wont to depict his opponents as weak, emasculated men: he presented not only Cicero, as we have seen, as ‘loose and unmanly’ (solutus et enervis ), but also the principal rulers of the city. 163 Calvus’ crusade against lasciviousness is as much directed against the in his view bombastic oratorical style as against what he considers to be degenerate conduct. For Calvus, then, literary style and life-style must follow the same basic rules: both speech and behavior should be restrained and masculine. One remaining question is: did Calvus uphold these moral standards in his own life? While otherwise very little is known about the historical Calvus, the elder Pliny gives us an interesting snippet of evidence in a discussion about the medicinal use of lead plates: ‘With these plates the orator Calvus is reported to have restrained himself and to have preserved his body’s strength for the labor of his studies.’ 164 At night, such sheets of lead were typically laid on the area of the loins and the

160 The reference in schol. Iuv. 9.133 to the completeness of the poem should not be taken for granted: see Jocelyn (1996). The poem has received its share of scholarly attention: see, e.g., Courtney (1993) 210, Jocelyn (1996) and Hollis (2007) 83–84. Jocelyn offers a thorough reading of the epigram, examining in great detail its textual problems, the scholarly tradition and its cultural context: he concludes that the verses comment on Pompey’s ‘desire for an active male sexual partner’. Zanker (2016) 61 thinks that the principal pun in the poem is derived from the ambiguity of sibi velle : ‘The point of Calvus’ epigram lay in the fact that sibi velle can be interpreted as both ‘to want for oneself’ and ‘to mean’.’ The connection between Pompey’s strange tic and homosexuality is widely referred to in Antiquity: see, e.g., Sen. Ep. 52.12, Juv. 9.133, and Plut. Mor. 89e, 800d, Pomp. 48, Caes. 4, and Amm. Marc. 17.11.4. 161 Plut. Mor. 89e: Βάδισμα τρυφερώτερον εἰς μαλακίαν διέβαλε , καὶ Πομπήιον τὸ ἑνὶ κνᾶσθαι τὴν κεφαλὴν δακτύλῳ πορρωτάτω θηλύτητος καὶ ἀκολασίας ὄντα . 162 On this connotation of the word vir , see Jocelyn (1996) 254. 163 See Tac. Dial. 18.4–5, quoted in section 5.6.1 above. 164 Plin. NH 34.166: His lamnis Calvus orator cohibuisse se traditur viresque corporis studiorum labori custodisse. While Pliny himself already connects Calvus’ self-therapy to the ‘labor of his studies’ ( studiorum labor ), Dugan (2001) 401–6 further explores the relationship between Calvus’ use of lead and his stylistic

245

CHAPTER FIVE kidneys, in order to check sexual disorders, such as involuntary erections and spontaneous discharges of semen, which in ancient medicine were thought to make men ‘shriveled, weak, high-voiced, hairless, beardless and effeminate’ ( ῥικνοί , ἀσθενέες , ὀξύφωνοι , ἄτριχες , ἀγένειοι , γυναικώδεες ). 165 Thus, Pliny’s testimony suggests that Calvus used lead to control his sexual urges, to restrain himself and to preserve his manliness. Likewise, it has been suggested that his lost work On the Use of Cold Water , known only from an enigmatic reference in Martial, may have discussed the anaphrodisiac properties of cold water. 166 Although the anecdote in Pliny may be apocryphal, it is consistent with Calvus’ other activities: his mantras of self-restraint and masculinity are linked to all aspects of his legacy— to his rhetorical theory, to his poetics, to his approach to politics, to his moral views and even to the stories about his personal life. In fact, his purism is already apparent in his two cognomina—‘meager’ (macer ) and ‘bald’ ( calvus )—corresponding not only to his slender physical appearance but also marvelously appropriate for the Atticist (life-)style that he so passionately advertised. Thus, in Calvus’ case it is undeniable that style is the man himself.

5.7 Conclusion This chapter has been concerned with the overwhelming admiration for Attic oratory that pervades the extant Greek and Latin sources on prose style of the first century BC. Following recent scholarship on Atticism, which has rejected the misguided thesis that there existed a discourse, which, as we have seen (section 5.6.1 above), often refers to the imagery of the human body: ‘The body discourse within these domains of rhetorical theory follows the same logic as Calvus’ medical program, thus demonstrating the vitality of these metaphors within rhetoric, and revealing continuities between rhetorical and medical conceptions of the body.’ 165 Aretaeus, On the Causes and Signs of Chronic Diseases 4.5. On the basis of the medical volumes of Aristotle, Aretaeus, Galen and Priscian, Dugan (2001) 403 notes that lead plates were specifically prescribed for curing ‘satyriasis (an insatiable desire for sexual activity), priapism (a chronic desireless erection), and the involuntary discharge of semen that the Greeks called gonorrhea (in Latin, seminis lapsus or seminis effusio )’. 166 See Mart. 14.196: ‘These pages that tell you of fountains and the names of rivers were better swimming in their own water’ ( haec tibi quae fontes et aquarum nomina dicit / ipsa suas melius charta natabat aquas ). The title of Martial’s epigram is ‘Calvus’ On the Use of Cold Water ’ ( Calvi de aquae frigidae usu ). The joke appears to be that the treatise on the applications of cold water is worthless to the extent that it might just as well be tossed back into the very water, whence the book (that is, the papyrus) came. That it is indeed our Calvus who is the author of the work is suggested by the fact that the foregoing epigram (14.195) is entitled ‘Catullus’, after his friend. Isetta (1977) argues that On the Use of Cold Water is the ultimate source of Pliny’s reference to Calvus’ use of lead plates. Her argument is supported by Dugan (2001) 406: ‘The linkage between cooling therapies and the control of wet dreams supports Sandra Isetta’s suggestion.’

246

TO BE ATTIC OR NOT TO BE ATTIC uniform school promoting the revival of Attic prose style, I have argued that there was in fact more than one Attic muse in Rome: Calvus, Cicero and Dionysius could each confidently present their own stylistic programs, despite the conspicuous differences between them, as thoroughly Attic. Thus, our authors each painted a different picture of their ideal form of eloquence: Calvus depicts his muse as a natural beauty who does not need any feminine jewelry or make-up, Cicero imagines her to be a damsel in distress who must be saved by all mean necessary from ruthless villains, and Dionysius presents her as a faithful housewife who judiciously holds sway over her domestic domain. These interpretations of Attic style are all built on a shared conception of the defining virtues of Athenian oratory, politics and moral character: while Classical Athens was consistently associated with such attributes as purity, vigor, wisdom, democracy, restraint and masculinity, the authors in Rome each highlighted different elements from this Athenian cultural repertoire not only in accordance with their aesthetic tastes, but also in order to suit their views on contemporary society. During the turbulent transition from Republic to Empire, Athens and Attica could be used as flexible buzzwords to demarcate one’s moral, political and civic views in stylistic terms: after all, an author’s style was regarded as a window onto his very soul. Thus, the extant discussions of prose style offer us interesting perspectives on famous episodes in the history of Rome, such as the first triumvirate, Caesar’s dictatorship and the early reign of Augustus. We have seen, for instance, that Calvus’ simple and subdued Atticism resonates with his crusade against extravagance and effeminacy, for which he denounces the triumvirs Caesar and Pompey. Furthermore, I have observed that Cicero’s grand and vigorous Atticism accommodates his vehement struggle for the preservation of the republican constitution, which was threatened by Caesar’s violent policies. The moderation and balance of Dionysius’ middle-style Atticism, lastly, suits the Greek teacher’s conviction that the city of Rome has brought about a renaissance of the civic values that once formed the basis of Athens’ glory: the emperor himself, by all appearances, seems to have agreed with this optimistic assessment. In my project of recovering the goals and motivations that underlie the stylistic discourse of Late-Republican and Augustan Rome, this chapter has shown how closely the theories of style are connected to the authors’ views of politics, society and life.

247

Chapter 6 GENERAL CONCLUSION

In this dissertation, I have offered, for the first time in more than a century, a systematic comparison of the two most important extant experts on prose style from the first century BC —the Roman orator, rhetorician, politician and philosopher Cicero, and the Greek historian, literary critic and teacher of rhetoric Dionysius of Halicarnassus. While previous discussions of their relationship were primarily concerned with identifying the sources of their common ideas, I have adopted a synchronic approach to their works: I have consulted Cicero and Dionysius (as well as several of their fellow critics and rhetoricians) on how to compose great prose in Late-Republican and Augustan Rome. My attention, in other words, has been focused on the Greek and Roman discourse on stylistic theory during the city’s turbulent transition from Republic to Empire. In this final chapter, I will briefly summarize the most important conclusions of this study. Each of the foregoing chapters concentrates on a separate recurring theme in the stylistic discussions of Cicero, Dionysius and their Greek and Roman colleagues— specifically, the enthusiastic praise for the Athenian orator Demosthenes and the ruthless criticism of the Hellenistic orator and historian Hegesias of Magnesia (chapter 2), the division of style into three basic types as a means to analyze, categorize and compare artistic prose (chapter 3), the emphasis on the aural aesthetics of literature through the careful review of the harmonic and rhythmical aspects of word arrangement (chapter 4), and the exploitation of Atticism as a malleable vehicle for promoting one’s ideas about brilliant eloquence, prudent politics and superior morality (chapter 5). These chapters not only discuss the Greek and Latin sources for these specific topics, but they also exemplify and corroborate my three overarching contentions about stylistic theory in Late-Republican and Augustan Rome (section 1.1).

First, the various surviving stylistic discussions draw on and contribute to a shared discourse on prose style: the obvious connections between the exposés of Cicero, Dionysius and other (near-)contemporary authors can thus be attributed to their familiarity with a common framework of theoretical concepts, analytical tools and technical vocabulary. The existence of such a widely available repertoire can be ascertained, for instance, in the oft-attested threefold history of rhetoric, distinguishing between a golden age in Classical Athens, a subsequent

249

CHAPTER SIX period of decline in Hellenistic Asia, and an ultimate era of revival in contemporary Rome (section 1.6): on the basis of this classicizing periodization, Cicero and Dionysius adopt a prejudiced approach to the Classical prose of the Athenian Demosthenes and to the Hellenistic prose of the Asian-born Hegesias (section 2.2). Moreover, in their evaluations of the stylistic achievements of these two opposite men, Cicero and Dionysius reach similar conclusions: they both applaud Demosthenes’ stylistic versatility, the force of his thunderbolts, and his virtuoso word arrangement (section 2.3), while they criticize Hegesias on exactly these points, viz., his formulaic monotony, his lack of vigor, and the awkwardness of his rhythmical composition (section 2.4). These like-minded analyses of genius and ineptitude reveal Cicero’s and Dionysius’ joint approach to the evaluation of prose style. In other aspects of their stylistic theories, too, the authors apply a shared critical apparatus, which stood at the disposal of Greek and Roman writers alike. We have seen, for instance, that Cicero and Dionysius both build their three-style formulas on an antithesis between two diametrically opposed extreme styles, supplemented by a third intermediate type (section 3.2), while their classifications of Greek prose authors according to these three categories are largely compatible with each other (section 3.3). Concerning the theory of word arrangement, Cicero and Dionysius present a strikingly similar comparison between oratory and theatrical performance that serves to illustrate the same basic natural law of combination (section 4.3), and they also dictate analogous principles for the production of rough and smooth acoustics (section 4.4). As for the omnipresent Atticism of our sources, I have observed that Cicero, his opponent Calvus, and Dionysius articulate their conceptions of Attic style by selecting elements from a commonly accessible reservoir of political and moral virtues that were typically associated with Classical Athens (section 5.3): each author, hence, presents his Lady Eloquence as a pure, prudent, freeborn and moderate woman (section 5.4). In sum, there is more than enough common ground between Cicero, Dionysius and their fellow scholars to vouch for my contention that they participated in a close-knit common discourse.

The second main thesis of this dissertation holds that the elements from the universal theoretical and analytical toolbox are by no means rigid or monolithic, but rather flexible and fluid. Critics and rhetoricians could adopt (or not adopt) and adapt these elements in accordance with their own goals and motivations. Throughout my investigations I have established several reasons that may account for the various stylistic views within the extant corpus of sources. One decisive factor appears to have been the pursuit of ad hoc writing

250

GENERAL CONCLUSION purposes, as I have argued in my discussion of the theory of three styles: the emphasis on the simple and grand registers in Orator seems to be part of Cicero’s strategy to refute the claims of the self-styled Attic orators in Rome (section 3.4), whereas the focus on the intermediate register in On Demosthenes can be connected to Dionysius’ case for a versatile style with universal appeal (section 3.5). The fact that neither author consistently retains the three-style doctrine throughout his oeuvre should not be understood in the context of a development in their thought: rather, I have argued that Cicero and Dionysius use their critical tools for pragmatic reasons, omitting them from their discussions if they have no use for them. Another factor in the diverse applications of stylistic theory is aesthetic taste. Particularly, there appears to be a general difference between the Greek and Roman literary palate: Dionysius, for instance, assumes that the uneducated masses appreciate simple oratory, whereas the intellectual elite craves for an elevated style, but according to his Roman colleagues the situation is completely reversed (section 3.6). Greek crowds apparently require a different approach from the speaker than Roman audiences. Furthermore, Greek and Roman authors as a rule exhibit disparate attitudes toward verbal arrangement: the Romans tend to be suspicious of meticulously arranged words, although many Greeks are wont to praise the art of composition as the culmination of creative artistry (section 4.2). Accordingly, our sources preserve two approaches to the disharmonious tones and limping rhythms of rough arrangement: in Latin texts, they are usually regarded as emblematic of authentic sincerity (section 4.5.1), and in Greek discussions as signs of sublime grandeur (section 4.5.2). In other cases, however, we have seen that the borderline between Greece and Rome does not account for differences of taste: Roman Calvus and Greek Dionysius both praise crude, ear-piercing sounds as the time-honored echoes from a legendary past, while Cicero rejects such acoustic effects as the outdated remnants from a simplistic stone age (section 4.5.3). Additionally, the varied range of the extant stylistic opinions can be related to the political and moral views of the respective authors. Cicero’s conception of Attic style, for one, cannot be separated from his resistance to Caesar’s dictatorship: he resorts to powerful, passionate oratory as the only type of eloquence that he considers capable of protecting the republic against tyranny (section 5.5.1). Dionysius’ insistence on a well-balanced middle style, alternatively, reflects the civic virtues of Classical Athens, like justice and moderation, which in his view were being revived in contemporary Rome (section 5.5.2). Strikingly, Augustus’ propaganda closely parallels this approach: the emperor pushes an agenda of moral restoration, and he presents himself as a moderate man, whose oratory keeps a due mean. Calvus, finally, not only applies the ‘Attic’ principles of restraint and manliness to his stylistic

251

CHAPTER SIX theories but also to his other activities: he criticizes the leading politicians of his day, Caesar and Pompey, for their supposed extravagance and effeminacy (section 5.6.1), while he seemingly recommends the use of lead and cold water to prevent such vices in private life (section 5.6.2). In a word, the critics and rhetoricians variously interpret elements from their common discourse in order to demarcate their position in Roman society as authors, citizens and men.

My third crucial claim, which has been a continuous thread running through this dissertation, is that Greek and Roman authors alike engage in a mutual exchange of ideas on prose style. In my view, we should not strictly mark off separate Roman and Greek stylistic traditions nor should we impose an inherent hierarchy on the surviving Greek and Latin sources: all texts should be regarded as full-fledged contributions to a common discourse. I have proposed the notion of a ‘koine’, a buzzword in recent scholarship on cultural interaction, to make sense of the relationship between Greek and Roman scholars: all these authors, with their diverse interests and objectives, speak the same language, that is, a rhetorical-critical koine (section 1.7). This common, intercultural discourse functions on the basis of two interrelated processes: the theories, techniques and terminologies of individual participants can become part of the koine (‘universalization’), and can thus subsequently be used by other participants, who can adapt them to suit their own objectives (‘particularization’), whence they can once again end up in the koine, and so on. In my opinion, we thus have a convenient conceptual framework to understand how both Greeks and Romans could be active contributors to a shared discourse on prose style, whilst allowing us to interpret their views in the context of their own goals and motivations. From these considerations it can be inferred that Greek and Roman authors must have been familiar, directly or indirectly, with each other’s work. Here we are getting into murkier waters: Roman authors do not refer explicitly to the teachings of contemporary Greek colleagues, or vice versa. Furthermore, as the elements of the shared repertoire are often widely attested, it is difficult to establish who inspires whom: I have therefore been reluctant to identify instances of direct Ciceronian influence in Dionysius’ treatises, although Cicero’s overwhelming impact on the stylistic debate in Rome is incontestable. That there must have been an intimate interaction between Greek and Roman scholars can be seen in the debate about word arrangement: while Roman authors painstakingly distanced themselves from Greek approaches to composition, Dionysius in turn seems to address the concerns of his Roman colleagues in his treatise on the topic. In addition, the obsession with Attic style

252

GENERAL CONCLUSION appears not to have taken hold in Rome before the quarrel between the renowned Roman orators Calvus and Cicero, after which it inevitably became a commonplace topic for Greek scholars working in the city who subsequently passed their conceptions of the issue on to their Roman students (section 5.2). It is very likely, then, that Greek and Roman authors not only had access to a shared discourse, but that both groups also developed it further in a continuous dialogue.

This dissertation has proposed a threefold conceptual framework to understand the relationship between the stylistic theories of Cicero, Dionysius and contemporary authors from the first century BC: both the striking similarities and the dramatic differences between their views can be fruitfully explained by assuming that (1) the critics and rhetoricians participate in a commonly available stylistic discourse, that was (2) prone to flexible, pragmatic interpretation, and which was (3) exploited and dialectically elaborated further by Greek and Roman authors alike. In a nutshell, I submit that this study has cast new light on the mechanisms and motivations underlying the extant conceptions of great prose in the literary heyday that was Late-Republican and Augustan Rome: we have seen that the era’s approach to style is intimately bound up with the diverse literary, cultural and political ambitions of the various authors. According to popular wisdom, there is no accounting for tastes—and yet, we now have a plausible way of accounting for the literary tastes of Rome’s principal stylistic arbiters.

253

Bibliography

ACHARD , G. (1985) ‘L’auteur de la Rhétorique à Herennius ?’, Revue des Études Latines 64: 56–68.

ACHARD , G. (ed.) (1989) La Rhétorique à Herennius . Paris.

ADAMIETZ , J. (1960) Ciceros De Inventione und die Rhetorik ad Herennium. Marburg.

ADAMIK , T. (1995) ‘Cicero’s Theory of Three Kinds of Style’, Acta Classica Universitatis Scientiarum Debreceniensis 31: 3–10.

ADAMS , C.D. (1917) ‘Demosthenes’ Avoidance of Breves’, Classical Philology 12: 271–294.

ADAMS , J.N. (1982) The Latin Sexual Vocabulary. London.

ADAMS , J.N., M. JANSE and S. SWAIN (eds.) (2002) Bilingualism in Ancient Society: Language Contact and the Written Text. Oxford.

AHERN KNUDSEN , R. (2010) Homeric Speech and the Origins of Rhetoric. Baltimore.

ALBRECHT , M. VON (2003) Cicero’s Style: A Synopsis Followed by Selected Analytic Studies. Leiden and Boston.

ALLAN , R.J., I.J.F. DE JONG , C.C. DE JONGE (2017) ‘From Enargeia to Immersion: The Ancient Roots of a Modern Concept’, Style 51.1: 34–51.

ALLEN , W.S. (1978) Vox Latina: A Guide to the Pronunciation of Classical Latin . Cambridge. [second revised edition]

AMMON , G. (1889) De Dionysii Halicarnassensis librorum rhetoricum fontibus. Diss. München.

ANASTASSIOU , A.A. (1965) Zur antike Wertschätzung der Beredsamkeit des Demosthenes . Diss. Kiel.

ARMSTRONG , D. (1995a) ‘The Impossibility of Metathesis: Philodemus and Lucretius on Form and Content in Poetry’, in Obbink (1995): 210–232.

ARMSTRONG , D. (1995b) ‘Philodemus, On Poems Book 5’, in Obbink (1995): 255–269.

ARTÉS HERNÁNDEZ , J.A. (2014) ‘ Λέξις in Dionysius of Halicarnassus’ Writings on Rhetoric’, Rhetorica 31.4: 372–387.

ASMIS , E. (2004) ‘Philodemus on the Sound and Sense of in the Poetry’, Cronache Ercolanesi 34: 5–28.

AUBERT -BAILLOT , S. and C. GUÉRIN (eds.) (2014) Le Brutus de Cicéron: Rhétorique, politique et histoire culturelle . Leiden and Boston.

255

BIBLIOGRAPHY

AUGELLO , I. (2006) Cecilio di Calatte: Frammenti di critica letteraria, retorica e storiografia. Rome.

AUGUSTYNIAK , C. (1957) De tribus et quattuor dicendi generibus quid docuerint antiqui. Warsaw.

AUJAC , G. (1978) Denys d’Halicarnasse , Opuscules rhétoriques , Tome I: Les orateurs antiques . Paris.

AUJAC , G. (1988) Denys d’Halicarnasse , Opuscules rhétoriques , Tome II: Démosthène . Paris.

AUJAC , G. (1991) Denys d’Halicarnasse , Opuscules rhétoriques , Tome IV: Thucydide, Seconde Lettre à Ammée . Paris.

AUJAC , G. (1992) Denys d’Halicarnasse , Opuscules rhétoriques , Tome V: L'imitation (fragments, épitomé), Première lettre à Ammée, Lettre à Pompée Géminos, Dinarque . Paris.

AUJAC , G. and M. LEBEL (1981) Denys d’Halicarnasse , Opuscules Rhétoriques , Tome III: La composition stylistique . Paris.

AUSTIN , R.G. (1948) Quintiliani institutionis oratoriae liber XII. Oxford.

AX, W. (1978) ‘ Ψόφος , φωνή und διάλεκτος als Grundbegriffe aristotelischer Sprachreflexion’, Glotta 56: 245–271.

BABBITT , F.C. (ed.) (1936) Plutach, Moralia, Volume V: Isis and Osiris, The E at Delphi, The Oracles at Delphi No Longer Given in Verse, The Obsolescence of Oracles. Cambridge, MA.

BALSDON , J.P.V. (1971) ‘Dionysius on Romulus: A Political Pamphlet?’, Journal of Roman Studies 61: 18–27.

BARAZ , Y. (2012) A Written Republic: Cicero’s Philosophical Politics. Princeton and Oxford.

BARKER , A.D. (2014) ‘Greek Musical Theorists on the Sound of Speech’, Annali della Scuola Normale Superiore di Pisa, Classe di Lettere e Filosofia 6.2: 657–679.

BARWICK , K. (ed.) (1925) Artis grammaticae libri V . Leipzig.

BATTISTI , D.G. (ed.) (1997) Dionigi di Alicarnasso Sull’Imitazione, Edizione critica, traduzione e commento . Pisa and Rome.

BAYLISS , A.J. (2011) After Demosthenes: The Politics of Early Hellenistic Athens . London and New York.

BEAUJARD , P. (2010) ‘From Three Possible Iron-Age World Systems to a Single Afro- Eurasian World-System’, Journal of World History 21: 1–43.

BERRY , D.H. (1996) ‘The Value of Prose Rhythm in Questions of Authenticity: The Case of

256

BIBLIOGRAPHY

De Optimo Genere Oratorum attributed to Cicero’, Papers if the Leeds International Latin Seminar 9: 47–74.

BILDE , P.G. (1993) ‘The International Style: Aspects of Pompeian First Style and Its Eastern Equivalents’, in Aspects of Hellenism in Italy: Towards a Cultural Unity? , eds. P.G. Bilde, I. Nielsen and M. Nielsen. Copenhagen: 151–177.

BISHOP , C. (2011) Greek Scholarship and Interpretation in the Works of Cicero. Diss. Philadelphia.

BISHOP , C. (2016) ‘How to Make a Roman Demosthenes: Self-Fashioning in Cicero’s Brutus and Orator ’, The Classical Journal 111.2: 167–192.

BLAENSDORF , F. J. (1994) ‘Cratès et les débuts de la philologie romaine’, in Grammaire et rhétorique: Notion de romanité (Actes du colloque de Strasbourg, novembre 1990, Contributions et travaux de l’Institut d’histoire romaine de l’Université des Sciences humaines de Strasbourg , VII ), ed. J. Dangel. Strasbourg: 5–11.

BLASS , F. (1880) Die attische Beredsamkeit, 3. Abteilung, 2. Abschnitt: Demosthenes’ Genossen und Gegner . Leipzig.

BLASS , F. (1893) Die attische Beredsamkeit, 3. Abteilung, 1. Abschnitt: Demosthenes. Leipzig. [second revised edition]

BLASS , F. (1905) Die Rhythmen der asianischen und römischen Kunstprosa . Leipzig.

BONNER , S.F. (1938) ‘Dionysius of Halicarnassus and the Peripatetic Mean of Style’, Classical Philology 33: 257–266.

BONNER , S.F. (1939) The Literary Treatises of Dionysius of Halicarnassus: A Study in the Development of Critical Method. Cambridge.

BORNECQUE , H. (1907) Les clausules métriques latines. Lille.

BOTTAI , F. (1999a) ‘Aspetti della tecnica espositiva di Dionigi di Alicarnasso nel De oratoribus antiquis ’, Prometheus 25: 45–60.

BOTTAI , F. (1999b) ‘Aspetti della tecnica espositiva di Dionigi di Alicarnasso nel De oratoribus antiquis ’, Prometheus 25: 132–150.

BOWERSOCK , G.W. (1965) Augustus and the Greek World. Oxford.

BOWERSOCK , G.W. (1979) ‘Historical Problems in Late Republican and Augustan Classicism’, in Flashar (1979): 57–78.

BRINGMANN , K. (1971) Untersuchungen zum späten Cicero. Göttingen.

BRINK , C.O. (1963) Horace on Poetry: Prolegomena to the Literary Epistles. Cambridge.

BRINK , C.O. (1971) Horace on Poetry: The ‘Ars Poetica’. Cambridge.

BRINK , C.O. (1982) ‘Quintilian’s De causis corruptae eloquentiae and Tacitus’ Dialogus de

257

BIBLIOGRAPHY

oratoribus ’, Classical Quarterly 39: 472–503.

LA BUA , G. (2014) ‘Cicero’s Pro Milone and the ‘Demosthenic’ style: De optimo genere oratorum 10’, Greece and Rome 61.1: 29–37.

BURSTEIN , S.M. (1989) Agatharchides of Cnidus On the Erythraean Sea . London.

BURSTEIN , S.M. (2012) ‘Agatharchides of Knidos (86)’, in Brill’s New Jacoby , ed. I. Worthington. Published online at www.referenceworks.brillonline.com.

BUTLER , S. (2011) The Matter of the Page: Essays in Ancient and Medieval Authors. Madison, WI.

CALBOLI , G. (1969) Cornifici Rhetorica ad Herennium. Bologna.

CALBOLI , G. (1997a) ‘Asianismus’, in Der Neue Pauly: Enzyklopädie der Antike , eds. H. Cancik and H. Schneider. Stuttgart and Weimar: 79–80.

CALBOLI , G. (1997b) ‘Attizismus’, in Der Neue Pauly: Enzyklopädie der Antike , eds. H. Cancik and H. Schneider. Stuttgart and Weimar: 262–263.

CALBOLI , G. (2003) ‘Vipstanus Messalla in Tacitus’ Dialogus de oratoribus ’, in Petroniana: Gedenkschrift für Hubert Petersmann , eds. J. Herman, H. Rosén, H. Gärtner and H.A. Gärtner. Heidelberg: 67–82.

CALBOLI MONTEFUSCO , L. (1994) ‘Aristotle and Cicero on the Officia oratoris ’, in Peripatetic Rhetoric after Aristotle, eds. W.W. Fortenbaugh, D.C. Mirhady. New Brunswick, NJ and London: 66–94.

CALCANTE , C.M. (2000) Genera dicendi e retorica del sublime. Pisa and Rome.

CALCANTE , C.M. (2004) ‘Il De elocutione di Demetrio e il De Demosthene di Dionigi di Alicarnasso’, Rendiconti Istituto Lombardo 138: 99–124.

CALCANTE , C.M. (2005a) Eufonia e onomatopea: Interpretazioni dell'iconismo nell'antichità classica . Como.

CALCANTE , C.M. (2005b) ‘Da Cratilo a Dionigi d’Alicarnasso: Eufonia e iconismo’, Athenaeum 93.1: 5–50.

CAPLAN , H. (ed.) (1954) [Cicero] Rhetorica ad Herennium . Cambridge, MA.

CAPLAN , H. (1970) ‘The Decay of Eloquence at Rome in the First Century’, in Of Eloquence: Studies in Ancient and Medieval Rhetoric by Harry Caplan , eds. A. King and H. North. London and Ithaca, NY: 160–195.

CARTLEDGE , P.A. (1993) The Greeks: A Portrait of Self and Others. Oxford.

CARY , E. (ed.) (1937) The Roman Antiquities of Dionysius of Halicarnassus, Vol. 1. Cambridge, MA.

CASTORINA , E. (1946) Licinio Calvo. Catania.

258

BIBLIOGRAPHY

CHAHOUD , A. (2004) ‘The Roman Satirist Speaks Greek’, Classics Ireland : 1–46.

CHIRON , P. (ed.) (1993) Démétrios, Du style. Paris.

CHIRON , P. (2001) Un rhéteur méconnu: Démétrios (Ps.-Démétrios de Phalère): Essai sur les mutations de la théorie du style à l’époque hellénistique. Paris.

CHIRON , P. (2014) ‘Démétrios de Phalère dans le Brutus ’, in Aubert-Baillot and Guérin: 105–120.

CICU , L. (2000) ‘Cicerone e il prepon’, Paideia 55: 123–162.

CITRONI , M. (1998) ‘Percezioni di classicità nella letterature latina’, in Che cos’è il classicismo? , eds. R. Cardini and M. Regoliosi. Rome: 1–34.

CITRONI , M. (2003) ‘I canoni di autori antichi: Alle origini del concetto di classico’, in Culture europee e tradizione latine, Atti del Convegno internazionale di studi, Cividale del Friuli, 16 –17 novembre 2001 , eds. L. Casarsa, L. Cristante and M. Fernandelli. Trieste: 1–22.

CITRONI , M. (2006) ‘The Concept of the Classical and the Canons of Model Authors in Roman Literature’, in Porter (2006a): 204–234.

COLVIN , S. (2011) ‘The Koine: A New Language for a New World’, in Creating a Hellenistic World , eds. A. Erskine and L. Llewellyn-Jones. Swansea: 31–45.

CONNOLLY , J. (2007a) The State of Speech: Rhetoric and Political Thought in Ancient Rome. Princeton.

CONNOLLY , J. (2007b) ‘Virile Tongues: Rhetoric and Masculinity’, in A Companion to Roman Rhetoric , eds. W.J. Dominik and J. Hall. Oxford and Malden, MA: 83–97.

CONNORS , C. (1997) ‘Field and Forum: Culture and Agriculture in Roman Rhetoric’, in Roman Eloquence: Rhetoric in Society and Literature, ed. W.J. Dominik. London: 58–73

COOPER , C. (2000) ‘Philosophers, Politics, Academics: Demosthenes’ Rhetorical Reputation in Antiquity’, in Demosthenes: Statesman and Orator , ed. I. Worthington. London: 224–245.

COPELAND , R. (1991) Rhetoric, Translation, and Hermeneutics in the Middle Ages. Cambridge.

CORBEILL , A.P. (1996) Controlling Laughter: Political Humor in the Late Roman Republic. Princeton.

CORNISH , F.W., J.P. POSTGATE and J.W. MACKAIL (eds.) (1966) Catullus, Tibullus, Pervigilium Veneris . Cambridge, MA. [revised edition by G.P. Goold]

COSTIL , P. (1949) L’esthétique littéraire de Denys d’Halicarnasse . Diss. Paris.

259

BIBLIOGRAPHY

COURTNEY , E. (ed.) (1993) The Latin Fragmentary Poets. Oxford.

CRUTTWELL , C.T. (1877) A History of Roman Literature. Oxford.

DAMON , C. (1991) ‘Aesthetic Response and Technical Analysis in the Rhetorical Writings of Dionysius of Hallicarnassus’, Museum Helveticum 48: 33–58.

DA RIOS , R. (ed) (1954) Aristoxeni Elementa Harmonica. Rome.

DELARUE , F. (1982) ‘L’asianisme à Rome’, Revue des Études Latines 60: 166–185.

DELCOURT , A. (2005) Lecture des Antiquités Romaines de Denys d’Halicarnasse: Un historien entre deux mondes . Brussels.

DENEIRE , T. (2004) ‘A Textbook in Latin Prose Composition? Virtutes and Vitia dicendi in Rhet. Her. 4,12–16’, in Virtutes Imago: Studies on the Conceptualisation and Transformation of an Ancient Ideal , eds. G. Partoens, G. Roskam and T. van Houdt. Louvain, Namur, Paris and Dudley, MA: 89–115.

DENCH , E. (2017) ‘Ethnicity, Culture, and Identity’, in The Oxford Handbook of the Second Sophistic , eds. D.S. Richter and W.A. Johnson. New York: 99–114.

DENNISTON , J.D. (1952) Greek Prose Style. Oxford.

DEROUX , C. (2008) ‘Petite histoire cocasse d’un mot coquin: ‘Salaputium, -ii’’, in Linguista sum: Mélanges offertes à Marc Dominic à l’occasion de son soixantième anniversaire , ed. E. Danblon. Paris: 133–146.

DIHLE , A. (1955) ‘Ein Spurium unter den Werken Ciceros’, Hermes 83: 303–314.

DIHLE , A. (1957) ‘Analogie und Attizismus’, Hermes 85: 170–205.

DIHLE , A. (1977) ‘Der Beginn des Attizismus’, Antike und Abendland 23: 162–177.

DIHLE , A. (1992) ‘Attizismus’, in Historisches Wörterbuch der Rhetorik, Band 1: A –Bib , ed. G. Ueding. Darmstadt: 1163–1176.

DONADI , F. (1999) ‘Il bello e il piacere (osservazioni sul De compositione verborum di Dionigi d’Alicarnasso)’, Studi Italiani di Filologia Classica 4.1: 42–63

DONADI , F. (2000a) ‘Il caso di Egesia: Sulla possibilità e convenienza di una nuova edizione del De compositione verborum di Dionigi d’Alicarnasso’, in Letteratura e riflessione sulla letteratura nella cultura classica , ed. G. Arrighetti. Pisa: 327–343.

DONADI , F. (2000b) Lettura del De compositione verborum di Diogini d’Alicarnasso . Padua.

DONADI , F. and A. MARCHIORI (2013) Dionigi d’Alicarnasso: La Compositione Stilistica . Trieste.

DOUGLAS , A.E. (1955) ‘M. Calidius and the Atticists’, Classical Quarterly 49: 241–247.

DOUGLAS , A.E. (1956) ‘Cicero, Quintilian, and the Canon of Ten Attic Orators’, Mnemosyne 9: 30–40.

260

BIBLIOGRAPHY

DOUGLAS , A.E. (1957) ‘A Ciceronian Contribution to Rhetorical Theory’, Eranos 55: 18–26.

DOUGLAS , A.E. (1958) Review of Portalupi (1957), Classical Review 8.4: 286.

DOUGLAS , A.E. (1966) M. Tulli Ciceronis Brutus . Oxford.

DOUGLAS , A.E. (1973) ‘The Intellectual Background to Cicero’s Rhetorica: A Study in Method’, Aufstieg und Niedergang der römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forshung, Joseph Vogt zu seinem 75. Geburtstag gewidmet, I.3 , ed. H. Temporini-Vitzthum. Berlin: 95–138.

DOVER , K.J. (1997) The Evolution of Greek Prose Style. Oxford.

DRERUP , E. (1923) Demosthenes im Urteile des Altertums. Würzburg.

DUECK , D. (2017) ‘Introduction’, in The Routledge Companion to Strabo , ed. D. Dueck. Abingdon, UK and New York: 1–7.

DUGAN , J. (1998) Review of Narducci (1997), Bryn Mawr Classical Review 1998.01.27.

DUGAN , J. (2001) ‘Preventing Ciceronianism: C. Licinius Calvus’ Regimens for Sexual and Oratorical Self-Mastery’, Classical Philology 96: 550–572.

DUGAN , J. (2005) Making a New Man: Ciceronian Self-Fashioning in the Rhetorical Works. Oxford.

DUGAN , J. (2007) ‘Modern Critical Approaches to Roman Rhetoric’, in A Companion to Roman Rhetoric , eds. W.J. Dominik and J. Hall. Oxford and Malden: 9–22.

DUGAN , J. (2012) ‘ Scriptum and Voluntas in Cicero’s Brutus ’, in Letteratura e Civitas: Transizioni dalla Repubblica all’Impero, In ricordo di Emanuele Narducci , ed. M. Citroni. Pisa: 117–128.

DUGAN , J. (2013) ‘Cicero’s Rhetorical Theory’, in Steel (2013): 25–40.

DYCK , A.R. (2008) ‘Rivals into Partners: Hortensius and Cicero’, Historia 57.2: 142–173.

EDER , W. (2005) ‘Augustus and the Power of Tradition’, in The Cambridge Companion to the Age of Augustus , ed. K. Galinksy. Cambridge: 13–32.

EGGER , M. (1902) Denys d’Halicarnasse . Paris.

ENGELS , J. (2004) ‘Agatharchides von Knidos’ Schrift Über das Rote Meer’, in Ad Fontes!: Festschrift für Gerhard Dobesch zum 65. Geburtstag am 15. September 2004 , dargebracht von Kollegen, Schülern und Freunden , eds. H. Heftner and K. Tomaschitz. Vienna: 179–192.

FALASCHI , E. (2017) ‘Looking at the Bronze of Lost Sculptures: The Reception of the Delphic Monument of the Admirals in the Imperial Age’, in Artistry in Bronze: The Greeks and Their Legacy (XIXth International Congress on Ancient Bronzes), eds. J. Daehner, K.D.S. Lapatin and A. Spinelli. Los Angeles: 94–97.

261

BIBLIOGRAPHY

FALCONER , W.A. (ed.) (1923) Cicero: On Old Age, On Friendship, On Divination. Cambridge, MA.

FANTHAM , E. (1979) ‘On the Use of Genus -Terminology in Cicero’s Rhetorical Works’, Hermes 107: 441–459.

FANTHAM , E. (2004) The Roman World of Cicero’s ‘De oratore’. Oxford.

FERRARA , G. (1970) ‘Commenti al dopoguerra aziaco’, La Cultura 8: 22–39.

FLASHAR , H. (ed.) (1979) Le classicisme à Rome aux 1ers siècles avant et après J.-C. Vandoeuvres, Geneva.

FOCKE , F. (1923) ‘Synkrisis’, Hermes 58: 327–368.

FORDYCE , C.J. (ed.) (1961) Catullus. Oxford.

FORMARIER , M. (2011) ‘L’orateur romain chantait-il?’, Synergies Espagne 4: 25–33.

FORMARIER , M. (2013) ‘ Ῥυθμός , rhythmos et numerus chez Cicéron et Quintilien: Perspectives esthétiques et génériques sur le rhythme oratoire latin’ in Rhetorica 31.2: 133–149.

FORNARO , S. (1997) Dionisio di Alicarnasso: Epistola a Pompeo Gemino, Introduzione e commento. Stuttgart and Leipzig.

FORNARO , S. (2001) ‘ Patina d’antico da Dionisio d’Alicarnasso a Winckelmann’, Sandalion 22: 35–45.

FORSDYKE , S. (2001) ‘Athenian Democratic Ideology and Herodotus’ Histories’, American Journal of Philology : 329–358.

FORTENBAUGH , W.W. (ed.) (1992) Theophrastus of Eresus: Sources for His Life, Writings, Thought and Influence, Volume 2: Psychology, Human Physiology, Living Creature, Botany, Ethics, Religion, Politics, Rhetoric, Poetics, Music, Miscellanea. Leiden.

FORTENBAUGH , W.W. (2005) Theophrastus of Eresus: Sources for His Life, Writings, Thought and Influence, Commentary Volume 8: Sources on Rhetoric and Poetics (Texts 666 –713). Leiden and Boston.

FOX , M. (1996) Roman Historical Myths: The Regal Period in Augustan Literature. Oxford.

FOX , M. (2007) Cicero’s Philosophy of History. Oxford.

FOX , M. (2019) ‘The Prehistory of the Roman polis in Dionysius’, in De Jonge and Hunter (2019a): 180–200.

FREUDENBURG , K. (1993) The Walking Muse: Horace on the Theory of Satire . Princeton.

FROMENTIN , V. (1998) Denys d’Halicarnasse, Antiquités romaines 1: Introduction générale et Livre I. Paris.

FUHRER , T. (2003) ‘Was ist gute Dichtung? Horaz und der poetologische Diskurs seiner

262

BIBLIOGRAPHY

Zeit’, Rheinisches Museum 146: 346–364.

GABBA , E. (1960) ‘Studi su Dionigi di Alicarnasso I: La costituzione di Romolo’, Athenaeum 38: 175–225.

GABBA , E. (1982a) ‘La “storia di Roma arcaica” di Dionigi d’Alicarnasso’, Aufstieg und Niedergang der römischen Welt II 30.1: 799–816.

GABBA , E. (1982b) ‘Political and Cultural Aspects of the Classicistic Revival in the Augustan Age’, La Critica d’Arte 1: 43–65.

GABBA , E. (1991) Dionysius and The History of Archaic Rome . Berkeley, Los Angeles and Oxford.

GAINES , R.N. (2007) ‘Roman Rhetorical Handbooks’, in A Companion to Roman Rhetoric , eds. W. Dominik and J. Hall. Oxford: 163–180.

GALINKSY , K. (1981) ‘Augustus’ Legislation on Morality and Marriage’, Philologus 125: 126–144.

GALLO , L. (2011) ‘Appunti per un riesame di Agatharchide di Cnido’, Hormos 3: 68–76.

GARCEA , A. (ed.) (1996) Caesar’s De Analogia: Edition, Translation, and Commentary . Oxford and New York.

GARCEA , A. and V. LOMANTO (2014) ‘Hortensius dans le Brutus : Une polémique rhétorique sous forme d’éloge funèbre’, in Aubert-Baillot and Guérin (2014): 141–160.

GARNSEY , P. (1988) Famine and Food Supply in the Greco-Roman World: Responses to Risk and Crisis . Cambridge.

GEIGENMÜLLER , P. (1908) Quaestiones dionysianae de vocabulis artis criticae. Leipzig.

GELZER , T. (1979) ‘Klassizismus, Attizismus und Asianismus’, in Flashar (1979): 1–55.

GENTILI , B. (1990) ‘Il De compositione verborum di Dionigi di Alicarnasso: Parola, metro e ritmo nella comunicazione letteraria’, Quaderni Urbinati di Cultura Classica 36.3: 7–21.

GIBSON , C.A. (2002) Interpreting a Classic: Demosthenes and his Ancient Commentators . Berkeley and London.

GIORDANO , L. (2000) ‘Ottaviano Augusto, scrittore: Le lettere private’, Memorie dell’ Accademia delle Scienze di Torino 24.2: 3–52.

GLEASON , M.W. (1995) Making Men: Sophists and Self-Presentation in Ancient Rome. Princeton.

GOOLD , G.P. (1961) ‘A Greek Professorial Circle at Rome’, Transactions of the American Philological Association 92: 168–192.

263

BIBLIOGRAPHY

GÖTZ , G. and F. SCHÖLL (eds.) (1910) M. Terenti Varronis De Lingua Latina quae supersunt. Leipzig.

GOUDRIAAN , K. (1989) Over classicisme: Dionysius van Halicarnassus en zijn program van welsprekendheid, cultuur en politiek . Diss. Amsterdam.

GRIMALDI , M. (2004) ‘A proposito dell’ esegesi retorico-grammaticale: Qualche esempio di metafrasi / riscrittura’, Linguistica e Letteratura 29: 21–49.

GROOT , A.W. de (1919a) A Handbook of Ancient Prose-Rhythm. Groningen and The Hague.

GROOT , A.W. de (1919b) De numero oratorio Latino . Groningen and The Hague.

GROOT , A.W. de (1921) Der antike Prosarhythmus I . Groningen and The Hague.

GRUBE , G.M.A. (1952) ‘Thrasymachus, Theophrastus, and Dionysius of Halicarnassus’, American Journal of Philology 73: 251–267.

GRUBE , G.M.A. (1961) A Greek Critic: Demetrius on Style . Toronto.

GRUBE , G.M.A. (1965) The Greek and Roman Critics. London.

GRUBE , G.M.A. (1974) ‘Greek Historians and Greek Critics’, Phoenix 28: 73–80.

GRUEN , E.S. (1967) ‘Cicero and Licinius Calvus’, Harvard Studies in Classical Philology 71: 215–233.

GRUEN , E.S. (1971) ‘Some Criminal Trials of the Late Republic: Political and Prosopographical Problems’, Athenaeum 49: 54–69.

GRUEN , E.S. (1992) Culture and National Identity in Republican Rome. Ithaca, NY.

GRUEN , E.S. (2011) Rethinking the Other in Antiquity . Princeton and Oxford.

GUÉRIN , C. (2014) ‘ Oratorum bonorum duo genera sunt : La définition de l’excellence stylistique et ses conséquences théoriques dans le Brutus ’, in Aubert-Baillot and Guérin (2014): 161–189.

GUNDERSON , E. (2000) Staging Masculinity: The Rhetoric of Performance in the Roman World. Ann Arbor.

GUTIÉRREZ GONZÁLEZ , R. (2012) ‘Augustus on Style and Language: A Reassessment of Some Fragmenta and Testimonia’, in Papers on Rhetoric 11 , ed. L. Calboli Montefusco. Perugia: 110–133.

GUTZWILLER , K. (1969) Ψυχρός und ὄγκος : Untersuchungen zur rhetorischen Terminologie. Diss. Zürich.

HALL , E. (1989) Inventing the Barbarian: Greek Self-Definition through Tragedy. Oxford.

HALL , J.M. (2002) Hellenicity: between Ethnicity and Culture . Chicago.

HALLIWELL , S., W.H. FYFE , D.A. RUSSELL , D. INNES and W.R. ROBERTS (eds.) (1995)

Aristotle: Poetics; Longinus: On the Sublime; Demetrius: On Style . Cambridge, MA.

264

BIBLIOGRAPHY

HAMMERSTAEDT , J. (1992) ‘Der Schlussteil von Philodems drittem Buch über Rhetorik’, Cronache Ercolanesi 22: 9–117.

HARTOG , F. (1980) Le miroir d’Hérodote: Essai sur la répresentation de l’autre. Paris.

HÄUSSLER , R. (1995) ‘Zur Datierung der Schrift vom Erhabenen’, in Prinzipat und Kultur im 1. und 2. Jahrhundert: Wissenschaftliche Tagung der Friedrich-Schiller-Universität Jena und der Iwane-Dshawachischwili-Universität Tbilissi, 27. –30. Oktober 1992 in Jena , eds. B. Kühnert, V. Riedel and R.V. Gordeziani. Bonn: 141–163.

HEATH , M. (1999) ‘Longinus, On Sublimity ’ in Proceedings of the Cambridge Philological Society 45: 43–74.

HEATH , M. (2003) ‘‘Pseudo-Dionysius’ Art of Rhetoric 8–11: Figured Speech, Declamation, and Criticism’, American Journal of Philology 124: 81–105.

HECK , H. (1917) Zur Entstehung des retorischen Attizismus . Diss. München.

HELDMANN , K. (1979) ‘ Hic primus inflexit orationem und die gute alte Redekunst’, Rheinisches Museum 122: 317–325.

HELDMANN , K. (1982) Antike Theorien über Entwicklung und Verfall der Redekunst. Munich.

HENDERSON , J. (ed.) (1998) Aristophanes: Acharnians, Knights . Cambridge, MA.

HENDRICKSON , G.L. (1904) ‘The Peripatetic Mean of Style and the Three Stylistic Characters’, American Journal of Philology 25: 125–146.

HENDRICKSON , G.L. (1905) ‘The Origin and Meaning of the Ancient Characters of Style’, American Journal of Philology 26: 249–290.

HENDRICKSON , G.L. (1906) ‘The De Analogia of Julius Caesar: Its Occasion, Nature and Date, with Additional Fragments’, Classical Philology 1.2: 97–120.

HENDRICKSON , G.L. (1926) ‘Cicero’s Correspondence with Brutus and Calvus on Oratorical Style’, American Journal of Philology 13: 72–80.

HENDRICKSON , G.L. and H. HUBBELL (eds.) (1962) Cicero : Brutus, Orator . Cambridge, MA. [revised edition]

HERBERT -BROWN , G. (1999) ‘Jerome’s Dates for Gaius Lucilius, Satyrarum Scriptor ’ Classical Quarterly 49.2: 535–543.

HIDBER , T. (1996) Das klassizistische Manifest des Dionys von Halikarnass: Die praefatio zu De oratoribus veteribus . Stuttgart and Leipzig.

HILL , H. (1961) ‘Dionysius of Halicarnassus and the Origins of Rome’, Journal of Roman Studies 51: 88–93.

HOLLIS , A.S. (2007) Fragments of Roman Poetry c. 60 BC – 20 AD. New York.

265

BIBLIOGRAPHY

VAN HOOK , L. (1905) The Metaphorical Terminology of Greek Rhetoric and Literary Criticism . Chicago.

VAN HOOK , L. (1917) ‘ Ψυχρότης ἢ τὸ ψυχρόν ’, Classical Philology 12.1: 68–76.

HUBBELL , H.M. (1913) The Influence of Isocrates on Cicero, Dionysius and Aristides . Diss. New Haven and London.

HUBBELL , H.M. (1920) ‘The Rhetorica of Philodemus’, in Transactions of the Connecticut Academy of Arts and Sciences 23: 243–382.

HUBBELL , H.M. (ed.) (1949) Cicero : De inventione, De optimo genere oratorum, Topica. London and Cambridge, MA.

HUELSENBECK , B. (2018) Figures in the Shadows: The Speech of Two Augustan-Age Declaimers, Arellius Fuscus and Papirius Fabianus. Berlin and Boston.

HUNTER , R. (2009) Critical Moments in Classical Literature: Studies in the Ancient View of Literature and its Uses . Cambridge.

HURST , A. (1982) ‘Un critique grec dans la Rome d’Auguste: Denys d’Halicarnasse’, Aufstieg und Niedergang der römischen Welt II 30.1: 839–865.

HUTCHINSON , G.O. (1995) ‘Rhythm, Style, and Meaning in Cicero’s Prose’, Classical Quarterly 45.2: 485–499.

HUTCHINSON , G.O. (2013) Greek to Latin: Frameworks and Contexts for Intertextuality. Oxford.

HUTCHINSON , G.O. (2015) ‘Appian the Artist: Rhythmic Prose and its Literary Implications’, Classical Quarterly 65.2: 788–806.

HUTTON , M. and W. PETERSON (eds.) (1970) Tacitus: Agricola, Germania, Dialogue on Oratory. Cambridge, MA. [revised edition by R.M. Ogilvie, E.H. Warmington and M. Winterbottom]

INNES , D.C. (1985) ‘Theophrastus and the Theory of Style’, in Theophrastus of Eresus: on His Life and Work , eds. W.W. Fortenbaugh, P.M. Huby and A.A. Long. New Brunswick: 251–67.

INNES , D.C. (1989) ‘Augustan Critics’, in The Cambridge History of Literary Criticism, Vol. I: Classical Criticism , ed. G.A. Kennedy. Cambridge: 245–273.

INNES , D.C. (1994) ‘Period and Colon: Theory and Example in Demetrius and Longinus’, in Peripatetic Rhetoric after Aristotle , eds. W.W. Fortenbaugh and D.C. Mirhady. New Brunswick: 36–54.

INNES , D.C. (2002) ‘Longinus and Caecilius: Models of the Sublime’, Mnemosyne 55.3: 259–284.

266

BIBLIOGRAPHY

INNES , D.C. (2007) ‘The Written and Performative Styles’, in Influences on Peripatetic Rhetoric: Essays in Honor of William W. Fortenbaugh , ed. D.C. Mirhady. Leiden: 151–168.

DELL ' INNOCENTI PIERINI (1979) ‘Cicerone ‘demiurgo’ dell’oratore ideale: Riflessioni nel margine ad Orator 7–10’, Studi Italiani di Filologia Classica 51: 84–102.

IODICE DI MARTINO , M.G. (1986) ‘La metafora del corpo nelle opere retoriche di Cicerone’, Bollettino di studi latini 16: 22–30.

ISETTA , S. (1977) ‘Sul De Aquae Frigidae Usu di Calvo’, Studi e Ricerche dell’Istituto di Latino 1: 107–112.

JACOBY , F. (ed.) (1929) Die Fragmente der griechischen Historiker . Teil 2: Zeitgeschichte, B: Spezialgeschichten, Autobiographien und Memoiren, Zeittafeln [Nr. 106–261]. Berlin.

JACOBY , F. (ed.) (1930) Die Fragmente der griechischen Historiker. Teil 2: Zeitgeschichte, D: Kommentar zu Nr. 106 –261. Berlin.

JANKO , R. (ed.) (2000) Philodemus, On Poems, Book 1: Introduction, Translation and Commentary. Oxford.

JANKO , R. (ed.) (2011) Philodemus, On Poems, Books 3 –4, with the Fragments of Aristotle, On Poets. Oxford and New York.

JOCELYN , H.D. (1996) ‘C. Licinius Calvus, fr. 19 Büchner’, Eikasmos 7: 243–254.

JOHNSON , W.R. (2007) ‘Neoteric Poets’, in A Companion to Catullus , ed. M.B. Skinner. Malden, MA: 175–189.

JONES , W.H.S. (ed.) (1923) Hippocrates: Ancient Medicine; Airs, Waters, Places; Epidemics 1 and 3; The Oath; Precepts; Nutriment. Cambridge, MA.

DE JONGE , C.C. (2001) ‘ Natura artis magistra : Ancient Rhetoricians, Grammarians and Philosophers on Natural Word Order’, in Linguistics in the Netherlands 2001 (AVT Publications 18), eds. T. van der Wouden and H. Broekhuis. Amsterdam: 159–166.

DE JONGE , C.C. (2005) ‘Dionysius of Halicarnassus and the Method of Metathesis’, Classical Quarterly 55.2: 463–480.

DE JONGE , C.C. (2008) Between Grammar and Rhetoric: Dionysius of Halicarnassus on Language, Linguistics and Literature . Leiden.

DE JONGE , C.C. (2009) Review of Marini (2007), Bryn Mawr Classical Review 2009.08.12.

DE JONGE , C.C. (2012a) ‘Dionysius and Longinus on the Sublime: Rhetoric and Religious Language’, American Journal of Philology 133.2: 271–300.

DE JONGE , C.C. (2012b) Review of Wiater (2011), Bryn Mawr Classical Review 2012.06.41.

267

BIBLIOGRAPHY

DE JONGE , C.C. (2013) ‘Longinus 36.3: The Faulty Colossus and Plato’s Phaedrus’, Trends in Classics 5: 318–340.

DE JONGE , C.C. (2014a) ‘The Attic Muse and the Asian Harlot: Classicizing Allegories in Dionysius and Longinus’, in Valuing the Past in the Greco-Roman World. Proceedings from the Penn-Leiden Colloquia on Ancient Values VII , eds. J. Ker and C. Pieper. Leiden: 388–409.

DE JONGE , C.C. (2014b) ‘Ancient Theories of Style (Léxis)’, in Encyclopedia of Ancient Greek Language and Linguistics, Vol. 3: P-Z, Index, eds. G.K. Giannakis, et al. Leiden and Boston: 326–331.

DE JONGE , C.C. (2016) ‘Building Texts: Architecture, Music and the Teaching of Prose Composition’, Beiträge zur Geschichte der Sprachwissenschaft 26: 49–68.

DE JONGE , C.C. (2017) ‘Dionysius of Halicarnassus on Thucydides’, in The Oxford Handbook of Thucydides , eds. S. Forsdyke, E. Foster and R. Balot. Oxford: 641–658.

DE JONGE , C.C. (2018a) ‘Eumaeus, Evander and Augustus: Dionysius and Virgil on Noble Simplicity’, in Homer and the Good Ruler in Antiquity and Beyond , eds. J. Klooster and B. van den Berg. Leiden and Boston: 157–181.

DE JONGE , C.C. (2018b) ‘Demosthenes versus Cicero: Intercultural Competition in Ancient Literary Criticism’, in Competition in Classical Antiquity. Proceedings from the Penn- Leiden Colloquia on Ancient Values IX , eds. C. Damon and C. Pieper. Leiden and Boston: 300–323.

DE JONGE , C.C. (2019) ‘Dionysius and Horace: Composition in Augustan Rome’, in De Jonge and Hunter (2019a): 242–266.

DE JONGE , C.C. and R. HUNTER (eds.) (2019a) Dionysius of Halicarnassus and Augustan Rome: Rhetoric, Criticism and Historiography. Cambridge.

DE JONGE , C.C. and R. HUNTER (2019b) ‘Introduction’, in De Jonge and Hunter (2019a): 1–33.

KAIBEL , G. (1885) ‘Dionysios von Halikarnass und die zweite Sophistik’, Hermes 20: 497–513.

KEITH , A.M. (1999) ‘Slender Verse: Roman Elegy and Ancient Rhetorical Theory’, Mnemosyne 52.1: 41–62.

KENNEDY , D.F. (1992) ‘Augustan and Anti-Augustan: Reflections on Terms of Reference’, in Roman Poetry and Propaganda in the Age of Augustus, ed. A. Powell. London: 26–58.

KENNEDY , G.A. (1956) ‘Theophrastus and Stylistic Distinctions’, Harvard Studies in Classical Philology 62: 93–104.

268

BIBLIOGRAPHY

KENNEDY , G.A. (1957a) ‘The Ancient Dispute over Rhetoric in Homer’, American Journal of Philology 78.1: 23–35.

KENNEDY , G.A. (1957b) ‘Theophrastus and Stylistic Distinctions’, Harvard Studies in Classical Philology 62: 93–104.

KENNEDY , G.A. (1963) The Art of Persuasion in Greece. London.

KENNEDY , G.A. (1972) The Art of Rhetoric in the Roman World, 300 B.C. – A.D. 300. Princeton.

KENNEDY , G.A. (1989) ‘The Evolution of a Theory of Artistic Prose’, in The Cambridge History of Literary Criticism, Vol. I: Classical Criticism, ed. G.A. Kennedy. Cambridge: 184–199.

KENNEDY , G.A. (1991) Aristotle, On Rhetoric: A Theory of Civic Discourse. New York and Oxford.

KENNEDY , G.A. (1995) A New History of Classical Rhetoric. Princeton.

KENNEDY , G.A. (2003) Progymnasmata: Greek Textbooks of Prose Composition and Rhetoric. Atlanta.

KIM , L. (2014) ‘Archaizing and Classicism in the Literary Historical Thinking of Dionysius of Halicarnassus’, in Valuing the Past in the Greco-Roman World (Proceedings from the Penn-Leiden Colloquia on Ancient Values VII, Mnemosyne Supplements 369), eds. C. Pieper and J. Ker. Leiden: 357–387.

KIRCHNER , R. (2005) ‘Die Mysterien der Rhetorik: Zur Mysterienmetapher in rhetoriktheoretischen Texten’, Rheinisches Museum 148: 165–180.

KOHL , R. (1915) De scholasticarum declamationum argumentis ex historia petitis. Diss. Münster.

KOLLER , H. (1954) Die Mimesis in der Antike: Nachahmung, Darstellung, Ausdruck. Bern.

KOLLER , H. (1955) ‘Stoicheion’, Glotta 34: 161–174.

KOZNIEWSKI , D. (1968) Griechische Metrik . Darmstadt.

KRENKEL , W. (ed.) (1970) Lucilius, Satiren . Leiden.

KROLL , W. (1907) ‘Randbemerkungen’, Rheinisches Museum 62: 86–101.

KROLL , W. (ed.) (1913) M. Tulli Ciceronis Orator . Berlin.

KROLL , W. and B. KYTZLER (eds.) (1962) Cicero , Brutus . Leipzig.

LAURAND , L. (1907) Études sur le style des discours de Cicéron avec une esquisse de l’histoire du ‘cursus’ (3 vol.). Paris.

LAUSBERG , H. (2008) Handbuch der literarische Rhetorik: Eine Grundlegung der Literaturwissenschaft. Stuttgart. [revised edition]

269

BIBLIOGRAPHY

LEBEK , W.D. (1970) Verba prisca: Die Anfänge des Archaisierens in der lateinischen Beredsamkeit und Geschichtsschreibens. Göttingen.

LEBEL , M. (1973) ‘Evolution de la doctrine de Denys d’Halicarnasse, du De Lysia aux De Compositione Verborum et De Demosthene II ’, Cahiers d’Études Anciennes 2: 79–88.

LEEMAN , A.D. (1955) ‘Le genre et le style historique à Rome’, Revue des Études Latines 33: 183–208.

LEEMAN , A.D. (1963) Orationis Ratio: the Stylistic Theories and Practices of the Roman Orators, Historians, and Philosophers . Amsterdam.

LEEMAN , A.D., H. PINKSTER and J. WISSE (1996) M. Tullius Cicero, De oratore libri III, 4. Bd.: Buch II, 291 –367, Buch III, 1 –95. Heidelberg.

LEEMAN , A.D., H. PINKSTER , J. WISSE , M. WINTERBOTTOM and E. FANTHAM (2008) M. Tullius Cicero, De oratore libri III, Vol. 5: A Commentary on Book III, 96 –230. Heidelberg.

LEHMANN , A. (2004) ‘Les débuts de la critique littéraire à Rome’, in Naissance de la science dans l’Italie antique et moderne: Actes du colloque franco-italien des 1er et 2eme décemebre 2000 (Université de Haute-Alsace) , eds. L. de Poli and Y. Lehmann. Bern and Frankfurt am Main: 139–162.

LEIDL , C.G. (2003) ‘The Harlot’s Art: Metaphor and Literary Criticism’, in Metaphor, Allegory, and the Classical Tradition: Ancient Thought and Modern Revisions , ed. G.R. Boys-Stones. Oxford and New York: 31–54.

LEO , F. (1889) ‘Die beiden metrischen Systeme des Altertums’, Hermes 117: 43–62.

LIEBERG , G. (1956) ‘Der Begriff ‘Structura’ in der Lateinischen Literatur’, Hermes 84.4: 455–477.

LIVINGSTONE , N. (2007) ‘Writing Politics: Isocrates’ Rhetoric of Philosophy’, Rhetorica 25.1: 15–34.

LONG , A.A. (1995) ‘Cicero’s Plato and Aristotle’, in Cicero the Philosopher , ed. J.G.F. Powell. Oxford: 37–62.

LONGO AURICCHIO , F. (1977) ‘ Φιλοδήμου Περὶ ῥητορικῆς libros primum et secundum ’, in Ricerche sui Papiri Ercolanesi 3 , ed. F. Sbordone. Napels: 2–277.

LÓPEZ MOREDA , S. (2003) ‘Sobre la polisemia de elegantia : De Plauto a Frontón’, Revista de Estudios Latinos 3: 45–69.

LOSSAU , M.J. (1964) Untersuchungen zur antiken Demosthenesexegese. Bad Homburg von der Höhe.

270

BIBLIOGRAPHY

LUCARINI , C.M. (2015) ‘I due stile asiani (Cic. Brut. 325, P. Artemid.) e l’origine dell’Atticismo letterario’, Zeitschrift für Papyrologie und Epigraphik 193: 11–24.

LURAGHI , N. (2003) ‘Dionysios von Halikarnassos zwischen Griechen und Römern’, in Formen römischer Geschichtsschreibung von den Anfängen bis Livius: Gattungen – Autoren – Kontexte , eds. U. Eigler, U. Gotter, N. Luraghi and U. Walter. Darmstadt: 268–286.

LUZZATTO , M.T. (1988) ‘L’oratoria, la retorica e la critica letteraria dalle origini ad Ermogene’, in Da Omero agli Alessandrini: Problemi e figure della letterature greca , ed. F. Montanari. Rome: 207–257.

MAC DONALD , C. (1977) Cicero, In Catilinam 1 –4, Pro Murena, Pro Sulla, Pro Flacco. Cambrdige, MA.

MAC PHAIL , E.M. (2014) Dancing around the Well: The Circulation of Commonplaces in Renaissance Humanism. Leiden and Boston.

MAIER , F.K. (2018) ‘Wahrheitlichkeit im Sinne der enargeia : Geographie und Geschichte bei Agatharchides’, in Die symphonischen Schwestern: Narrative Konstruktion von ‘Wahrheiten’ in der nachklassischen Geschichtsschreibung , eds. T. Blank and F.K. Maier. Stuttgart: 209–225.

MALCOVATI , H. (ed.) (1953) Oratorum Romanorum fragmenta liberae rei publicae. Turin.

MANGONI , C. (ed.) (1993) Filodemo, Il quinto libro della Poetca (PHerc. 1425 e 1538) . Naples.

MANKIN , J. (2011) Cicero , De Oratore: Book III. Cambridge.

MANSFELD , J. (1994) Prolegomena: Questions to be Settled Before the Study of an Author or a Text. Leiden and New York.

MARCHETTI , C.C. (2009) Aristoxenus’ Elements of Rhythm: Text, Translation and Commentary with a Translation and Commentary on POxy 2687. Diss. New Brunswick.

MARCOTTE , D. (2001) ‘Structure et caractère de l'œuvre historique d'Agatharchide’, Historia 50.4: 385–435.

MARENGHI , G. (1971) Dionisio di Alicarnasso: Dinarco. Milan.

MARINI , N. (2007) Demetrio, Lo Stile . Rome.

MARINONE , N. (2004) Cronologia ciceroniana . Rome. [revised edition]

MAROUZEAU , J. (1921) ‘Pour mieux comprendre les textes latins: Essai sur la distinction des styles’, Revue de Philologie, de Littérature et d’Histoire Anciennes 45: 149–193.

MAROUZEAU , J. (1935) Traité de stylistique appliquée au latin . Paris.

271

BIBLIOGRAPHY

MARTIN , P.M. (1971) ‘La propagande augstéenne dans les Antiquités Romaines de Denys d’Halicarnasse (Livre 1)’, Revue des Études Latines 49: 162–179.

MARTINHO , M. (2010) ‘Caractérisation et noms du style moyen selon Denys d’Halicarnasse’, in Les noms du style dans l’antiquité Greco-Latin , eds. P. Chiron and C. Lévy. Louvain: 201–220.

MATTHEWS , V.J. (ed.) (1996) Antimachus of Colophon: Text and Commentary. Leiden.

MAY , J.M. (ed.) (2002) Brill’s Companion to Cicero: Oratory and Rhetoric. Leiden.

MAY , J.M. and J. WISSE (2001) Cicero: On the Ideal Orator . Oxford and New York.

MAYER , A. (1910) Theophrasti περὶ λέξεως Fragmenta . Leipzig.

MAZZACANE , R. (1982) ‘Santra, VII’, Studi Noniani, VII : 189–224.

MAZZUCCHI , C.M. (2010) Dionisio Longino, Del sublime: Introduzione, testo critico, traduzione e commentario. Milan. [revised edition]

MCCABE , D.G. (1981) The Prose-Rhythm of Demosthenes . New York.

MEERWALDT , J.D. (1920) Studia ad generum dicendi historiam pertinentia: Pars I, De Dionysiana virtutum et generum dicendi doctrina. Diss. Amsterdam.

MOREAU , J. (1962) Aristote et son école. Paris.

MORETTI , G. (1990) Acutum dicendi genus: Brevitas, oscurità, sottigliezze e paradossi nelle tradizioni retoriche degli Stoici . Trento.

MÜLLER , C. (1855) Geographi Graeci Minores, Volume I. Paris.

MÜNZER , F. (1926) ‘Licinius 113’ in Paulys Real-Encyclopädie der classischen Altertumswissenschaft XIII.1, ed. A.F. von Pauly, F. Wissowa, W. Kroll et al. Stuttgart: 428–435.

NARDUCCI , E. (1997) Cicerone e l’eloquenza romana: Retorica e progetto culturale. Rome and Bari.

NARDUCCI , E. (2002a) ‘ Brutus : the History of Roman Eloquence’, in May (2002): 401–425.

NARDUCCI , E. (2002b) ‘ Orator and the Definition of the Ideal Orator’, in May (2002): 427–433.

NASSAL , F. (1910) Aesthetisch-rhetorische Beziehungen zwischen Dionysius von Halicarnassus und Cicero. Diss. Tübingen.

NEWSOME , D.J. (2011) Review of Wallace-Hadrill (2008), Rosetta 9: 67–74.

NISBET , R.G.M. (1990) ‘Cola and clausulae in Cicero’s Speeches’, in Owls to Athens: Essays on Classical Subjects Presented to Sir Kenneth Dover , ed. E.M. Craik. Oxford: 349– 359.

272

BIBLIOGRAPHY

NORDEN , E. (1898) Die antike Kunstprosa vom VI. Jh. v. Chr. bis in die Zeit der Renaissance. Leipzig.

NORLIN , G. (ed.) (1928) Isocrates: To Demonicus, To Nicocles or The Cyprians, Panegyricus, To Philip, Archidamus . Cambridge, MA.

NORTH , H. (1966) Sophrosyne: Self-Knowledge and Self-Restraint in Greek Literature. Ithaca, NY.

NÜNLIST , R. (2009) The Ancient Critic at Work: Terms and Concepts of Literary Criticism in Greek Scholia. Cambridge.

OBBINK , D. (ed.) (1995) Philodemus and Poetry: Poetic Theory and Practice in Lucretius, Philodemus and Horace. New York and Oxford.

OBERHELMAN , S. and D. ARMSTRONG (1995) ‘Satire as Poetry and the Impossibility of Metathesis in Horace’s Satires ’, in Obbink (1995): 233–254.

OOMS , S. and C.C. DE JONGE (2013) ‘The Semantics of ἐναγώνιος in Greek Literary Criticism’, Classical Philology 108: 95–110.

OSBORNE , R. and C. VOUT (2010) ‘A Revolution in Roman History?’, Journal of Roman Studies 100: 233–245.

O’S ULLIVAN , N. (1994) Alcidamas, Aristophanes, and the Beginning of Greek Stylistic Theory. Stuttgart.

O’S ULLIVAN , N. (1997) ‘Caecilius, the ‘Canons’ of Writers, and the Origins of Atticism’, in Roman Eloquence: Rhetoric in Society and Literature, ed. W.J. Dominik. London: 32–49.

O’S ULLIVAN , N. (2015) ‘‘Rhetorical’ vs ‘Linguistic’ Atticism: A False Dichotomy?’, Rhetorica : 134–146.

PAVANO , G. (1958) Saggio su Tucidide: Introduzione, testo, traduzione, commento, appendice e indici. Palermo.

PAXIMADI , G. (1989) ‘Un capitolo di storia della linguistica antica tra Peripato e Stoa’, Maia 41: 223–228.

PAXSON , J.J. (1998) ‘Personification’s Gender’, Rhetorica 16.2: 149–179.

PEARSON , L. (ed.) (1990) Elementa Rhythmica: The Fragment of Book II and the Additional Evidence for Aristoxenean Rhythmic Theory. Oxford.

PEIRANO , I. (2010) ‘Hellenized Romans and Barbarized Greeks: Reading the End of Dionysius of Halicarnassus, “Antiquitates Romanae”’, Journal of Roman Studies 100: 32–53.

273

BIBLIOGRAPHY

PELLING , C. (2019) ‘Dionysius on Regime Change’, in De Jonge and Hunter (2019a): 203– 220.

PERNOT , L. (2006) L’ombre du tigre: Recherches sur la réception de Démosthène. Naples.

PETERSON , W. (1891) M. Fabi. Quintiliani institionis oratioriae liber decimus. Oxford.

PFEIFFER , R. (1968) History of Classical Scholarship: From the Beginnings to the End of the Hellenistic Age. Oxford.

PIOTROWICZ , L. (1915) De Hegesia Magnete rerum gestarum scriptore . Krakow.

POHL , K. (1968) Die Lehre von den drei Wortfügungsarten: Untersuchungen zu Dionysios von Halikarnass, De Compositione Verborum . Tübingen.

POHLENZ , M. (1924) ‘Eine politische Tendenzschrift aus Caesars Zeit’, Hermes 59: 157–189.

POHLENZ , M. (1933) Τὸ πρέπον : Ein Beitrag zur Geschichte des griechischen Geistes. Göttingen.

POLLITT , J.J. (1972) Art and Experience in Classical Greece. Cambridge.

POMEROY , S. (1997) Families in Classical and Hellenistic Greece: Representations and Realities . Oxford.

PORTALUPI , F. (1957) Sulla corrente rodiense . Turin.

PORTER , J.I. (1986) The Material Sublime: Towards a Reconstruction of Materialist Critical Discourse and Aesthetics in Antiquity. Diss. Berkeley.

PORTER , J.I. (1995) ‘O ἱ κριτικοί : A Reassessment’, in Greek Literary Theory after Aristotle: A Collection of Papers in Honour of D.M. Schenkeveld , eds. J.G.J. Abbenes, S.R. Slings and I. Sluiter. Amsterdam: 83–109.

PORTER , J.I. (2001) ‘Des son qu’on ne peut entendre: Cicéron, les ‘kritikoi’ et la tradition du sublime dans la critique littéraire’, in Cicéron et Philodème: La polémique en philosophie , eds. C. Auvray-Assayas and D. Delattre. Paris: 315–341.

PORTER , J.I. (ed.) (2006a) Classical Pasts: The Classical Traditions of Greece and Rome . Princeton and Oxford.

PORTER , J.I. (2006b) ‘What is ‘Classical’ about Classical Antiquity?’, in Porter (2006a): 1–65.

PORTER , J.I. (2006c) ‘Feeling Classical: Classicism and Ancient Literary Criticism’, in Porter (2006a): 301–352.

PORTER , J.I. (2010) The Origins of Aesthetic Thought in Ancient Greece: Matter, Sensation, and Experience . Cambridge.

PORTER , J.I. (2016) The Sublime in Antiquity. Cambridge.

POWELL , J.G.F. (2013a) ‘Cicero’s Style’ in Steel (2013): 41–72.

274

BIBLIOGRAPHY

POWELL , J.G.F. (2013b) ‘The Embassy of the Three Philosophers to Rome in 155 BC’, in Hellenistic Oratory: Continuity and Change , eds. C. Kremmydas and K. Tempest. Oxford: 219–247.

PRANDI , L. (2016) ‘Hegesias of Magnesia (142)’, in Brill’s New Jacoby , ed. I. Worthington. Published online at www.referenceworks.brillonline.com.

PREISSHOFEN , F. (1979) ‘Kunsttheorie und Kunstbetrachtung’, in Flashar (1979): 263–282.

PRITCHETT , W.K. (1975) Dionysius of Halicarnassus: On Thucydides. Berkeley, Los Angeles and London.

QUADLBAUER , F. (1958) ‘Die genera dicendi bis Plinius d. J.’, Wiener Studien 71: 55–111.

QUADLBAUER , F. (1962) Die antike Theorie der genera dicendi im lateinischen Mittelalter. Vienna.

RADEMAKER , A. (2007) Sophrosyne and the Rhetoric of Self-Restraint: Polysemy and Persuasive Use of an Ancient Greek Value Term. Diss. Leiden.

RADERMACHER , L. (1899a) ‘Studien zur Geschichte der antiken Rhetorik IV: Ueber die Anfänge des Atticismus’, Rheinisches Museum 54: 351–374.

RADERMACHER , L. (1899b) ‘Studien zur Geschichte der antiken Rhetorik V: Excurs, Theophrast περὶ λέξεως ’, Rheinisches Museum 54: 374–380.

RADERMACHER , L. (1910) ‘Hegesias 13’ in Paulys Real-Encyclopädie der classischen Altertumswissenschaft VII.2, ed. A.F. von Pauly, F. Wissows, W. Kroll et al. Stuttgart: 2607–2608.

RADERMACHER , L. (1951) Artium scriptores: Reste der voraristotelischer Rhetorik. Vienna.

RAMSEY , J.T. (ed.) (2003) Cicero, Philippics I-II. Cambridge.

REID , R.S. (1996) ‘Dionysius of Halicarnassus’s Theory of Compositional Style and the Theory of Literate Consciousness’, Rhetoric Review 15: 46–64.

RICHARDSON , J.S. (2012) Augustan Rome 44 BC to AD 14: The Restoration of the Republic and the Establishment of the Empire . Edinburgh.

RICHLIN , A. (1997) ‘Gender and Rhetoric: Producing Manhood in the Schools’, in Roman Eloquence: Rhetoric in Society and Literature , ed. W.J. Dominik. London and New York: 90–110.

RIGGSBY , A.M. (1991) ‘Elision and Hiatus in Latin Prose’, Classical Antiquity 10: 328–343.

RISPOLI , G.K. (1998) ‘L’ “errore necessario”: Per una poetica della δυσφωνία ’, Cronache Ercolanesi 28: 119–132.

ROBERTS , W.R. (1897) ‘Caecilius of Calacte: a Contribution to the History of Greek Literary Criticism’, American Journal of Philology 18.3: 302–312.

275

BIBLIOGRAPHY

ROBERTS , W.R. (1900) ‘The Literary Circle of Dionysius of Halicarnassus’, Classical Review 14: 439–442.

ROBERTS , W.R. (1901) Dionysius of Halicarnassus: The Three Literary Letters . Cambridge.

ROBERTS , W.R. (1902) Demetrius On Style: The Greek Text of Demetrius De elocutione. Cambridge.

ROBERTS , W.R. (1910) Dionysius of Halicarnassus: On Literary Composition. Cambridge.

ROCCONI , E. (2010) ‘Musica e retorica nel De compositione uerborum di Dionigi di Alicarnasso: Per un’ ipotesi sulle fonti ritmiche del trattato dionisiano’, Incontri Triestini di Filologia Classica 8: 175–190.

ROHDE , E. (1876) Die griechische Roman und seine Vorläufer. Leipzig.

ROHDE , E. (1886) ‘Die asianische Rhetorik und die zweite Sophistik’, Rheinisches Museum 41: 170–190.

ROISMAN , J. (2005) The Rhetoric of Manhood. Berkeley.

ROLFE , J.C. (ed.) (1927) Gellius: Attic Nights, Volume II: Books 6 –13. Cambridge, MA.

ROLFE , J.C. (ed.) 1946) Gellius: Attic Nights, Volum I: Books 1 –5. Cambridge, MA. [revised edition]

ROLFE , J.C. (ed.) (1998) Suetonius: Lives of the Caesars, Volume I: The Deified Julius, The Deified Augustus, Tiberius, Gaius Caligula. Cambridge, MA. [revised edition by D.W. Hurley]

RONCONI , F. (1998) ‘ De optimo genere oratorum : Storio di un abbozzo’, in Appunti romani di filologia: raccolta di studi e comunicazioni di filologia, linguistica e letteratura greca e latina, ed. A. Masaracchia. Rome: 43–68.

RONNET , G. (1952) Denys d’Halicarnasse, Sur le style de Démosthène: Introduction, traduction et commentaire. Diss. Paris.

ROSILLO -LÓPEZ , C. (2017) Public Opinion and Politics in the Late Roman Republic. Cambridge.

ROUSE , W.H.D. (ed.) (1975) Lucretius: On the Nature of Things . Cambridge, MA. [revised edition by M.F. Smith]

RUDD , N. (ed.) (2004) Horace: Odes and Epodes . Cambridge, MA.

RUIJGH , C.J. (1987) ‘ Μακρὰ τελεία et μακρὰ ἄλογος (Denys d’Halicarnasse, De la composition des mots, cap. 17 et 20): Le prolongement de la durée d’une syllabe finale dans le rhythme du mot grec’, Mnemosyne 40: 31–352.

RUSSELL , D.A. (1964) ‘Longinus’, On the Sublime. Oxford.

ROLFE , J.C. (1979) ‘Classicizing Rhetoric and Criticism: The Pseudo-Dionysian Exetasis

276

BIBLIOGRAPHY

and Mistakes in Declamation ’, in Flashar (1979): 113–134.

ROLFE , J.C. (ed.) (2002) Quintilian: The Orator’s Education, Volume IV: Books 9 –10. Cambridge, MA.

RUTHERFORD , I. (1998) Canons of Style in the Antonine Age: Idea –Theory and its Literary Context. Oxford.

SAID , E.W. (1978) Orientalism. New York.

SCAGLIONE , A. (1972) The Classical Theory of Composition. From its Origins to the Present: A Historical Survey. Chapel Hill.

SCHENKEVELD , D.M. (1964) Studies in Demetrius On Style. Amsterdam.

SCHENKEVELD , D.M. (1968) ‘O ἱ κριτικοί in Philodemus’, Mnemosyne 21: 176–215.

SCHENKEVELD , D.M. (1975a) ‘Theories of Evaluation in the Rhetorical Works of Dionysius of Halicarnassus’, Museum Philologum Londiniense 1: 93–107.

SCHENKEVELD , D.M. (1975b) Review of Scaglione (1972), Mnemosyne 28.2: 215–217.

SCHENKEVELD , D.M. (1983) ‘Linguistic Theories in the Rhetorical Works of Dionysius of Halicarnassus’, Glotta 61: 67–94.

SCHENKEVELD , D.M. (1988) ‘ Iudicia vulgi : Cicero, De oratore 3.195ff. and Brutus 183ff.’, Rhetorica 6: 291–305.

SCHENKEVELD , D.M. (1990) ‘Studies in the History of Ancient Linguistics III: The Stoic τέχνη περὶ φωνῆς’, Mnemosyne 43: 86–108.

SCHENKEVELD , D.M. (2000) ‘The Intended Public of Demetrius' ‘On Style’: The Place of the Treatise in the Hellenistic Educational System’, Rhetorica 18.1: 29–48.

SCHIPPERS , A.M. (2019) Dionysius and Quintilian: Imitation and Emulation in Greek and Latin Literary Criticism . Diss. Leiden.

SCHMID , Walter. (1959) Über die klassische Theorie und Praxis des antiken Prosarhythmus . Wiesbaden.

SCHMID , Wilhelm (1887–1897) Der Atticismus in seinen Hauptvertretern (4 vol.). Stuttgart.

SCHMID , Wilhelm (1894) ‘Zur antiken Stillehre aus Anlass von Proklos’ Chrestomathie’, Rheinisches Museum 49: 133–149.

SCHULZ , V. (2014) Die Stimme in der antiken Rhetorik (Hypomnemata 194). Göttingen.

SCHULZE , C. (2019) ‘Ways of Killing Women: Dionysius and the Deaths of Horatia and Lucretia’, in De Jonge and Hunter (2019a): 161–179.

SHACKLETON BAILEY , D.R. (ed.) (1977) Cicero , Epistulae ad Familiares , Vol. 1: 62 –47 BC. Cambridge and New York.

277

BIBLIOGRAPHY

SHACKLETON BAILEY , D.R. (ed.) (1993) Martial: Epigrams, Volume III: Books 11 –14. Cambridge, MA.

SHACKLETON BAILEY , D.R. (ed.) (1999) Cicero: Letters to Atticus, Volume IV. Cambridge, MA.

SHACKLETON BAILEY , D.R. (ed.) (2001) Cicero: Letters to Friends, Volume II: Letters 114 – 280. Cambridge, MA.

SHIPLEY , F.W. (1911) ‘The Heroic Clausula in Cicero and Quintilian’, Classical Philology 6.4: 410–418.

SKIMINA , S. (1937) État actuel des études sur le rythme de la prose grecque. Cracow.

SLUITER , I. (1990) Ancient Grammar in Context: Contributions to the History of Ancient Linguistic Thought. Diss. Amsterdam.

SLUITER , I. (1994). Review of Wisse (1989), Mnemosyne 47.2: 251–256.

SLUITER , I. (1997) ‘Parapleromatic Lucubrations’, in New Approaches to Greek Particles , ed. A. Rijksbaron. Amsterdam: 233–246.

SLUITER , I. (1999) ‘Communication, Eloquence and Entertainment in Augustine’s De doctrina christiana’, in The Impact of Scripture in Early Christianity , eds. J. den Boeft and M.L. van Poll-van de Lisdonk. Leiden: 245–267.

SLUITER , I. (2005) ‘Homer in the Dining Room: An Ancient Rhetorical Interpretation of the Duel between Paris and Menelaus (Plut. Quaest. Conv. 9.13)’, CW 98: 379–396.

SLUITER , I. (2010) ‘Textual Therapy: On the Relationship between Medicine and Grammar in Galen’, Hippocrates and Medical Education: Selected Papers Read at the XIIth International Hippocrates Colloquium, Universiteit Leiden, 24 –26 August 2005 , ed. H.F.J. Horstmanshoff . Leiden: 25–52.

SMITH , C.F. (ed.) (1919) Thucydides: History of the Peloponnesian War, Volume I: Books 1 – 2. Cambridge, MA.

SMITH , R.M. (1995) ‘A New Look at the Canon of the Ten Attic Orators’, Mnemosyne 48.1: 66–79.

SMYTH , H.W. (1920) A Greek Grammar for Colleges. Cambridge, MA.

SPAWFORTH , A. (2012) Greece and the Augustan Cultural Revolution. Cambridge.

SPENGEL , L. (ed.) (1854) Rhetores Graeci , Vol. 2. Leipzig.

SPINA , L. (1989) ‘Il racconto di un racconto: Egesia di Magnesia in Dionigi d’Alicarnasso, de comp. verb. VI 18.25–27’, Vichiana 18: 333–340.

SPINA , L. (2004a) ‘Riscrivere Candaule’, Rhetorica 17: 111–136.

SPINA , L. (2004b) ‘Esercizi di stile: Riscritture e metamorfosi testuali, un’ introduzione’,

278

BIBLIOGRAPHY

Linguistica e Letteratura 29: 9–12.

STAAB , G. (2004) ‘Athenfreunde unter Verdacht: Der erste Asianist Hegesias aus Magnesia zwischen Rhetorik und Geschichtsschreibung’ Zeitschrift für Papyrologie und Epigraphik 148: 127–150.

STAFFORD , E.J. (1998) ‘Masculine Values, Feminine Forms: On the Gender of Personified Abstractions’, in Thinking Men: Masculinity and its Self-Representation in Classical Antiquity , eds. L. Foxhall and J. Salmon. London: 43–56.

STEEL , C.E.W. (1998) Review of Narducci (1997), The Classical Review 48.2: 499–500.

STEEL , C.E.W. (2007) ‘Lost Orators of Rome’, in A Companion to Roman Rhetoric , eds. W.J. Dominik and J. Hall. Oxford and Malden: 237–249.

STEEL , C.E.W. (ed.) (2013) The Cambridge Companion to Cicero. Cambridge.

STERNBACH , L. (1930) ‘Przyczynki do Hegezjasza z Magnezji / Contribution à l’étude d’Hégésias de Magnésie’, Bulletin de l’Academie Polonaise de Cracovie 14: 43–44.

STROH , W. (1982) ‘Die Nachahmung des Demosthenes in Ciceros Philippiken’, in Éloquence et rhétorique chez Cicéron: Sept exposés suivis de discussions , ed. W. Ludwig. Vandoeuvres, Geneva: 1–40.

STROUP , S.C. (2003) ‘ Adulta Virgo : The Personification of Textual Eloquence in Cicero’s Brutus ’, Materiali e Discussioni per l’Analisi dei Testi Classici 50: 115–140.

STROUP , S.C. (2007) ‘Greek Rhetoric Meets Rome: Expansion, Resistance, and Acculturation’, in A Companion to Roman Rhetoric , eds. W.J. Dominik and J. Hall. Oxford and Malden: 23–37.

STROUP , S.C. (2010) Catullus, Cicero, and a Society of Patrons. Cambridge.

STROUX , J. (1912) De Theophrasti virtutibus dicendi . Leipzig.

STURTEVANT , E.A. and R.G. KENT (1915) ‘Elision and Hiatus in Latin Prose and Verse’, Transactions of the American Philological Association 46: 129–155.

SUDHAUS , S. (1892–1896) Philodemi volumina rhetorica (3 volumes). Leipzig.

SUSEMIHL , F. (1892) Geschichte der griechischen Literatur in der Alexandrinerzeit, 2. Band. Leipzig.

SWAIN , S. (1996) Hellenism and Empire: Language, Classicism, and Power in the Greek World, AD 50 –250. Oxford.

SYME , R. (1939) The Roman Revolution. Oxford.

TEUFFEL , W.S. (1870) Geschichte der römischen Literatur. Leipzig.

TOLKIEHN , J. (1908) ‘Dionysios von Halikarnass und Caecilius von Kalakte’, Wochenschrift für klassische Philologie 25: 84–86.

279

BIBLIOGRAPHY

TRAGLIA , A. (1982) ‘Elementi stilistici nel De lingua latina di Varrone’, Annali della Scuola Normale Superiore di Pisa, Classe di Lettere e Filosofia 12: 481–511.

TRÉDÉ , M. (1992) Kairos: L’à propos et l’occasion: Le mot et la notion, d’Homère à la fin du IVe siècle avant J.-C. Paris.

TUPLIN , C. (2010) ‘Hermesianax of Colophon (691)’, in Brill’s New Jacoby , ed. I. Worthington. Published online at www.referenceworks.brillonline.com.

USENER , H. (1913) ‘Ein altes Lehrgebaüde der Philologie’, in Kleine Schriften II . Leipzig: 265–314.

USENER , H. and L. RADERMACHER (1899) Opuscula Rhetorica: Dionysii Halicarnasei quae exstant , Vol. V . Stuttgart and Leipzig.

USENER , H. and L. RADERMACHER (1904–1929) Opscula Rhetorica: Dionysii Halicarnasei quae exstant , Vol. VI. Stuttgart and Leipzig.

USHER , S. (ed.) (1974) Dionysius of Halicarnassus: Critical Essays, Volume I . London and Cambridge, MA.

USHER , S. (ed.) (1985) Dionysius of Halicarnassus: Critical Essays, Volume II . London and Cambridge, MA.

USHER , S. (2000) ‘ Sententiae in Cicero Orator 137–9 and Demosthenes De corona ’, Rhetorica 26.2: 99–111.

VAAHTERA , J. (1997) ‘Phonetics and Euphony in Dionysius of Halicarnassus’, Mnemosyne 50: 586–595.

VACHER , M.-C. (ed.) (1993) Suétone, Grammairiens et Rhéteurs. Paris.

VARDI , A.D. (1996) ‘ Diiudicatio locorum : Gellius and the History of a Mode in Ancient Comparative Criticism’, Classical Quarterly 46.2: 492–514.

VASALY , A. (2002) ‘Cicero’s Early Speeches’, in May (2002): 71–111.

VENINI , P. (1976) ‘Storia greca e attualità romana in Cicerone’, Rendiconti 110: 272–280.

VERSLUYS , M.J. (2013) ‘Material Culture and Identity in the Late Roman Republic (c. 200–c. 20)’, in A Companion to the Archaeology of the Roman Republic , ed. J.D. Evans. London: 429–440.

VERSLUYS , M.J. (2015) ‘Roman Visual Material Culture as Globalising Koine ’, in Globalisation and the Roman World: World History, Connectivity and Material Culture , eds. M. Pitts and M.J. Versluys. Cambridge: 141–174.

VEYNE , P. (1979) ‘L’hellénisation de Rome et la problematique des acculturations’, Diogène 196: 1–29.

280

BIBLIOGRAPHY

VIIDEBAUM , L. (2019) ‘Dionysius and Lysias’ Charm’, in De Jonge and Hunter (2019a): 106–124.

VILJAMAA , T. (2003) ‘ Colon and Comma : Dionysius of Halicarnassus on the Sentence Structure’, in Syntax in Antiquity , eds. P. Swiggers and A. Wouters. Leuven, Paris and Dudley: 56–100.

VLASSOPOULOS , K. (2007) ‘Beyond and Below the Polis: Networks, Associations and the Writing of Greek History’, Mediterranean Historical Review 22: 11–22.

VOIT , L. (1934) Δεινότης : Ein antiker Stilbegriff. Diss. Leipzig.

VOLLGRAFF , W. (1949) ‘Elementum’, Mnemosyne 2.4: 89–115.

WALLACE -HADRILL , A. (1989) ‘Rome’s Cultural Revolution’, Journal of Roman Studies 79: 157–164.

WALLACE -HADRILL , A. (1998) ‘To Be Roman, Go Greek: Thoughts on Hellenization at Rome’, Bulletin of the Institute of Classical Studies 42: 79–91.

WALLACE -HADRILL , A. (2008) Rome’s Cultural Revolution . Cambridge.

WARDLE , D. (2014) Suetonius, Life of Augustus . Oxford.

WARMINGTON , E.H. (ed.) (1938) Remains of Old Latin, Volume III: Lucilius, the Twelve Tables. Cambridge, MA.

WEAIRE , G. (2012) ‘How to Talk to a Roman Student: The Teacher’s Authority in Dionysius of Halicarnassus’ De compostione verborum ’, Illinois Classical Studies 35–36: 46–67.

WEHRLI , F. (1946) ‘Der erhabene und der schlichte Stil in der poetischen-rhetorischen Theorie der Antike’, in Phyllobolia für Peter von der Müll , eds. O. Gigon, K. Meuli, W. Theiler, F. Wehrli and B. Wyss. Basel: 9–34.

WEHRLI , F. (1967) Die Schule des Aristoteles: Texte und Kommentar, Band II: Aristoxenos . Basel. [second revised edition]

WEHRLI , F. (1968) Die Schule des Aristoteles: Texte und Kommentar, Band IV: Demetrios von Phaleron. Basel. [second revised edition]

WEHRLI , F. (1969) Die Schulde des Aristoteles: Texte und Kommentar, Band X: Hieronymos von Rhodos, Kritolaos und seine Schüler, Rückblick: Der Peripatos in vorchristlicher Zeit. Basel. [second revised edition]

WEIL , H. (1844) De l’ordre des mots dans les langues anciennes comparées aux langues modernes. Paris.

WEST , M.L. (1982) Greek Metre. Oxford.

WHITMARSH , T. (1998) Review of Hidber (1996), Bryn Mawr Classical Review 1998.08.07.

281

BIBLIOGRAPHY

WHITMARSH , T. (2001) Greek Literature and the Roman Empire: The Politics of Imitation . Oxford.

WHITMARSH , T. (2013a) Beyond the Second Sophistic: Adventures in Greek Postclassicism. Berkeley and London.

WHITMARSH , T. (2013b) ‘Resistance is Futile? Greek Literary Tactics in the Face of Rome’, in Les Grecs héritiers des Romains: Huit exposés suivis de discussions , ed. P. Derron. Geneva: 57–85.

WIATER , N. 2011. The Ideology of Classicism: Language, History, and Identity in Dionysius of Halicarnassus. Berlin and New York.

VON WILAMOWITZ -MOELLENDORFF , U. (1900) ‘Asianismus und Atticismus’, Hermes 35: 1–52.

WILKINSON , L.P. (1963) Golden Latin Artistry. Cambridge.

WILMANNS , A. (ed.) (1869) De M. Terenti Varronis Libris Grammaticis. Berlin.

WINKEL , L.C. (1979) ‘Some Remarks on the Date of the Rhetorica ad Herennium ’, Mnemosyne 32: 327–332.

WINTERBOTTOM , M. (1974) Review of Scaglione (1972), Classical Review 24.2: 299–300.

WINTERBOTTOM , M. (1989) ‘Cicero and the Middle Style’, in Studies in Latin Literature and its Tradition in Honour of C.O. Brink , eds. J. Diggle, J.B. Hall and H.D. Jocelyn. Cambridge: 125–131.

WINTERBOTTOM , M. (2011) ‘On Ancient Prose Rhythm: The Story of the Dichoreus’, in Culture in Pieces: Essays on Ancient Texts in Honour of Peter Parsons , eds. D. Obbink and R. Rutherford. Oxford: 262–276.

WISSE , J. (1989) Ethos and Pathos from Aristotle to Cicero. Amsterdam.

WISSE , J. (1995) ‘Greeks, Romans and the Rise of Atticism’, in Greek Literary Criticism after Aristotle: a Collection of Papers in Honour of D.M. Schenkeveld , ed. J.G.J. Abbenes, S.R. Slings and I. Sluiter. Amsterdam: 65–82.

WISSE , J. (1998) Review of Hidber (1996), Bryn Mawr Classical Review 1998.08.06.

WISSE , J. (2002) ‘The Intellectual Background of Cicero’s Rhetorical Works’, in Brill’s Companion to Cicero: Oratory and Rhetoric, ed. J.M. May. Leiden: 331–374.

WITCHER , R. (2000) ‘Globalisation and Roman Imperialism: Perspectives on Identities in Roman Italy’, in The Emergence of State Identities in Italy in the First Millennium BC , eds. E. Herring and K. Lomas. London: 213–225.

282

BIBLIOGRAPHY

WOLLNER , D. (1886) Die von der Beredsamkeit aus der Krieger- und Fechtersprache entlehnten bildlichen Wendungen in den rhetorischen Schriften des Cicero, Quintilian und Tacitus. Diss. Landau.

WOOLF , G. (1994) ‘Becoming Roman, Staying Greek: Culture, Identity and the Civilising Process in the Roman East’, Proceedings of the Cambridge Philological Society 40: 116–143.

WOOTEN , C.W. (1975) ‘Le développement du style asiatique pendant l’époque hellénistique’, Revue des Études Grecques 88: 94–104.

WOOTEN , C.W. (1977) ‘Cicero’s Reactions to Demosthenes: A Clarification’, Classical Journal 73: 37–43.

WOOTEN , C.W. (1983) Cicero’s Philippics and their Demosthenic Model: The Rhetoric of Crisis. Chapel Hill.

WOOTEN , C.W. (1987) Hermogenes’ On Types of Style . Chapel Hill and London.

WOOTEN , C.W. (1989) ‘Dionysius of Halicarnassus and Hermogenes on the Style of Demosthenes’, American Journal of Philology 105: 576–588.

WOOTEN , C.W. (1991) ‘Abruptness in Demetrius, Longinus, and Demosthenes’, American Journal of Philology 112.4: 493–505.

WOOTEN , C.W. (1997) ‘Cicero and Quintilian on the Style of Demosthenes’, Rhetorica 15.2: 177–192.

WORTHINGTON , I. (1994) ‘The Canon of the Ten Attic Orators’, in Persuasion: Greek Rhetoric in Action , ed. I. Worthington. London: 244–263.

WRAY , D.L. (2001) Catullus and the Poetics of Roman Manhood. Cambridge and New York.

VAN WYK CRONJÉ , J. (1986) Dionysius of Halicarnassus: De Demosthene, a Critical Appraisal of the Status Quaestionis . Hildesheim.

YUNIS , H. (2019) ‘Dionysius’ Demosthenes and Augustan Atticism’, in De Jonge and Hunter (2019a): 83–105.

ZANGARA , A. (2008) ‘Voir l’histoire: Théories anciennes du récit historique’, Anabases 7: 249–256.

ZANKER , A.T. (2016) Greek and Latin Expressions of Meaning: The Classical Origins of a Modern Metaphor. Munich.

ZANKER , P. (1988) The Power of Images in the Age of Augustus . Munich. [Transl. C.H. Beck]

ZETZEL , J.E.G. (ed.) (1995) Cicero, De Re Publica, Selections. Cambridge.

ZIELI ŃSKI , T. (1904) Das Clauselgesetz in Ciceros Reden. Leipzig.

283

Samenvatting

OVER HET SCHRIJVEN VAN GEWELDIG PROZA

CICERO , DIONYSIUS VAN HALICARNASSUS EN STIJLTHEORIE IN LAAT -REPUBLIKEINS EN AUGUSTEÏSCH ROME

Hoofdstuk 1 Inleiding: doelen en uitgangspunten Deze dissertatie presenteert een systematische vergelijking van de ideeën van twee antieke geleerden over prozastijl. De geleerden in kwestie zijn de Romeinse redenaar, filosoof en politicus Cicero (106-43 v. Chr.) en de in Rome actieve Griekse historicus, literaire criticus en leraar in de retorica Dionysius van Halicarnassus (ca. 60-na 8 v. Chr.). Er zijn goede redenen voor een onderzoek naar de ideeën van juist deze twee mannen over juist dit onderwerp. Cicero en Dionysius zijn met afstand onze belangrijkste bronnen voor stilistische theorieën in de eerste eeuw v. Chr., een bloeiperiode van (Latijns en Grieks) artistiek proza: men denke aan Caesars Over de oorlog in Gallië , Livius’ Sinds de stichting van de stad , Diodorus Siculus’ Historische bibliotheek en Strabo’s Geografie . Ook het tweetal zelf heeft zich op dit gebied bepaald niet onbetuigd gelaten: Cicero’s talloze redevoeringen, brieven en filosofische traktaten worden nog altijd geprezen om hun stilistisch vernuft, en Dionysius’ literaire aspiraties hebben vaste vorm gekregen in zijn Romeinse Oudheden , een twintigdelige geschiedenis van het vroege Rome. Daarmee kunnen de twee geleerden ons veel leren over de heersende esthetische denkbeelden uit een tijd die nog altijd te boek staat als een literaire gouden eeuw. Kortom: Cicero en Dionysius bieden ons een masterclass over het schrijven van geweldig proza. De vorige studie naar de opvattingen van de twee schrijvers kende een diachrone inslag: de Duitse classicus Franz Nassal was in 1910 vooral geïnteresseerd in de voorgeschiedenis van hun beider gedachten. Dit proefschrift hanteert daarentegen een synchrone benadering: aan de hand van de retorisch-kritische werken van Cicero, Dionysius en enkele tijdgenoten richt het zich op de reconstructie van het discours over prozastijl in laat- republikeins en Augusteïsch Rome. Daarbij kan het bogen op recente studies die Cicero’s retorica bezien in het licht van diens zelfpresentatie in de late republiek, en op onderzoeken die Dionysius’ historische en kritische werken relateren aan literaire en politieke trends in het Rome van Augustus. De belangrijkste bronteksten zijn Cicero’s werken over de ‘redenaar’

285

SAMENVATTING

(orator ), te weten zijn De oratore (55 v. Chr.), zijn Brutus , Orator en De optimo genere oratorum (alle uit 46 v. Chr.) alsmede het voorwoord bij Dionysius’ Over de oude redenaars en diens traktaten Over Demosthenes en Over woordschikking (alle na 30 v. Chr.). Uit vier detailstudies over centrale thema’s in deze werken (hoofdstuk 2 t/m 5) vloeien de drie hoofdstellingen van dit proefschrift voort:

1. Cicero en Dionysius baseren zich op soortgelijke theorieën, ze maken gebruik van soortgelijke analytische methoden en ze bedienen zich van een soortgelijk technisch vocabulaire. Aangezien verscheidene elementen uit dit gemeenschappelijke repertoire ook aan de dag worden gelegd door andere Romeinse retorici en Griekse critici in Rome, kunnen we spreken van een breed gedeeld Grieks-Romeins discours over prozastijl.

2. Het gezamenlijke denkkader over prozastijl is niet rigide, maar flexibel: vanuit min of meer dezelfde uitgangspunten formuleren Cicero, Dionysius en hun tijdgenoten dan ook uiteenlopende stilistische ideeën. De auteurs stellen de gemeenplaatsen van het gedeelde discours onder andere ten dienste van hun eigen schrijfdoelen en literaire smaak. Zij laten zich bovendien leiden door hun persoonlijke morele kompas en politieke agenda.

3. Het gemeenschappelijke, flexibele vertoog dat we aantreffen in de bronnen is een vruchtbare voedingsbodem voor een dialoog tussen Griekse en Romeinse geleerden over prozastijl. Het traditionele ‘hellenocentrische’ perspectief, waarin wordt uitgegaan van unilaterale Griekse invloed op Romeinen, moet worden vervangen ten gunste van een meer dialectisch model dat een wederzijdse uitwisseling van ideeën tussen Grieken en Romeinen postuleert.

In hoofdstuk 1 worden deze drie fundamentele beweringen nader uitgewerkt en onderbouwd. In het bijzonder wordt daarbij de aandacht gevestigde op de notie van een ‘algemeen beschikbaar repertoire’ oftewel een koine (κοινή ), een term die veel gebezigd wordt in recente literatuur over cultureel contact in de antieke wereld. In het geval van Griekse en Latijnse retorisch-kritische teksten bestaat de koine uit een gevarieerde verzamelpoel van theorieën, technieken en terminologieën, waartoe Cicero, Dionysius en hun collega’s toegang hadden: elk van de algemeen beschikbare elementen vindt zijn oorsprong in een specifieke context,

286

SAMENVATTING bijvoorbeeld Aristoteles’ Retorica , Hellenistische poëziekritiek of het straffe oordeel van de oude Cato, maar is vervolgens steeds breder omarmd (‘universalization’) en kan, losgeweekt van zijn oorspronkelijke context, door gebruikers opnieuw worden ingezet in andere contexten (‘particularization’), waarna de nieuwe particuliere varianten weer in de koine terecht kunnen komen, enzovoorts. Dit cyclische model is een aantrekkelijke manier om zowel de overeenkomsten als de verschillen tussen de hoofdrolspelers in de geschiedenis van de stijlleer te verklaren: het laat zien hoe zowel Grieken als Romeinen, allen vanuit hun eigen motivaties en ambities, een actieve bijdrage konden leveren aan een gemeenschappelijk, zich continu ontwikkelend discours over prozastijl.

Hoofdstuk 2 Voorbeelden uit het verre verleden: Demosthenes en Hegesias Eén van de grondslagen van zowel de werken van Cicero als die van Dionysius is hun ‘classicisme’. Beide auteurs verdelen de geschiedenis van de literatuur in drie tijdvakken (de zogenaamde ‘klassizistische Dreischritt’): er is eerst een ‘klassieke’ bloeitijd, vervolgens een duistere periode van verval, die begint rond de dood van Alexander de Grote (323 v. Chr.), en uiteindelijk een wedergeboorte van de oude luister in het heden. Deze periodisering wordt nadrukkelijk gekoppeld aan drie mediterrane regio’s: het hoogtij wordt geassocieerd met redenaars uit Attica, het schiereiland waarop Athene ligt, de neergang met auteurs uit Klein- Azië, en het herstel met eigentijdse schrijvers in Rome. Ook roepen Cicero en Dionysius allebei de redenaar Demosthenes (384-322 v. Chr.) uit tot de allerbeste Attische schrijver en bombarderen ze eensgezind de historicus en redenaar Hegesias van Magnesia (ca. 300 v. Chr.) tot de allerslechtste Aziatische schrijver. Hoofdstuk 2 gaat na waarom Cicero en Dionysius juist Demosthenes de hemel in prijzen en waarom ze juist Hegesias hartstochtelijk verketteren. Tevens worden de stilistische oordelen getoetst aan de hand van passages uit de werken van de twee voorbeeldauteurs. Op die manier wordt achterhaald wat volgens Cicero en Dionysius de onderscheidende kenmerken zijn van goed en slecht proza. Demosthenes en Hegesias fungeren in de stijlkritiek als stereotypische symbolen voor respectievelijk de deugden van Attica en de zonden van Azië: Demosthenes wordt vereenzelvigd met de vrijheidslievende, democratische inborst van zijn moederstad Athene, terwijl Hegesias wordt afgeschilderd als een schoolvoorbeeld van de veronderstelde slaafse, teugelloze en verwijfde volksaard van de Aziaten. Deze bevooroordeelde clichés laten onverlet dat Cicero en Dionysius de teksten van de twee oude schrijvers onderwerpen aan een serieuze analyse, die voortborduurt op het werk van Hellenistische critici. Demosthenes wordt

287

SAMENVATTING geprezen om (1) zijn veelzijdigheid, die hem in staat stelt om bij elke gelegenheid de juiste toon aan te slaan; hij wordt ook bewonderd om (2) de kracht van zijn ‘bliksemschichten’, die bovenal van pas komt bij urgente kwesties van leven en dood; en hij oogst grote lof vanwege (3) zijn feilloze talent voor ritmische woordschikking, waarmee hij zijn gehoor betovert. Opvallend genoeg wordt Hegesias aangevallen op exact dezelfde punten: hem wordt verweten dat hij (1) formulair en monotoon is, waardoor hij zijn stijl niet kan aanpassen aan de omstandigheden; er wordt ook beweerd dat hij (2) de kracht mist om indruk te maken op het moment suprême; en er wordt gesteld dat hij (3) een eerder lach- dan ontzagwekkende ritmische woordschikking tentoonspreidt. Cicero en Dionysius gebruiken Demosthenes en Hegesias dus als aansprekende voorbeelden uit het verre Griekse verleden om vorm te geven aan hun ideeën over het schrijven van geweldig proza in het Romeinse heden: het gaat hun daarbij kennelijk vooral om caleidoscopische veelzijdigheid, overweldigende kracht en betoverende woordplaatsing. Hier worden we de contouren van het gedeelde stilistische discours in Rome gewaar. In de volgende hoofdstukken wordt nader ingegaan op de verschillende praktische uitwerkingen die Cicero, Dionysius en hun collega’s geven aan hun gezamenlijke denkraam, te beginnen met hun opvattingen over veelzijdig proza.

Hoofdstuk 3 De theorie van de drie stijlen Om verslag te kunnen doen van de veelzijdigheid die zij zo bewonderen in de teksten van Demosthenes (en de Attische schrijvers in het algemeen) maken meerdere auteurs in Rome onderscheid tussen drie stilistische grondvormen of ‘stijltypen’ (χαρακτῆρες τῆς λέξεως , genera dicendi ), grofweg een simpele stijl, een middenstijl en een verheven stijl. Een dergelijke driedeling vinden we in de eerste eeuw v. Chr. niet alleen bij Cicero en Dionysius, maar ook bij Varro (116-27 v. Chr.) en de anoniem overgeleverde Retorica voor Herennius (ca. 85-80 v. Chr.). Hoofdstuk 3 stelt dat de theorie van de drie stijlen een flexibel analytisch instrument is: geen twee auteurs gebruiken dezelfde terminologie in hun discussies over het onderwerp en elke auteur classificeert de klassieke Attische schrijvers telkens op zijn eigen manier volgens de drie categorieën. Dionysius beschikt zelfs over niet één maar twee stilistische driedelingen, één gericht op woordkeuze, één op woordschikking. Cicero en Dionysius zetten de drie stijlen op ingenieuze wijze naar hun hand. In zijn Orator kent Cicero aan elk stijltype een specifieke retorische functie toe: de simpele stijl dient om aan te tonen (probare ), de middenstijl om te vermaken ( delectare ) en de verheven stijl om

288

SAMENVATTING het publiek te manipuleren (flectere ). Op deze manier kan Cicero de driestijlenleer gebruiken in zijn felle polemiek tegen een prominente oratorische beweging (ca. 55-45 v. Chr.) die een kale stijl zonder franje (à la Lysias) propageerde en aanstoot nam aan Cicero’s weelderige stilistische ornamenten: omdat deze redenaars hun eigen sobere stijl beschouwden als de enige waarlijk Attische stijl, noemden zij zichzelf Attici en maakten zij Cicero uit voor ‘Aziaat’. Die laatste bijt van zich af door te stellen dat er niet één maar drie stijltypen bestaan, die alle drie door de redenaar beheerst moeten worden om de drie cruciale retorische functies te vervullen die van hem verlangd worden. Volgens Cicero kan de simpele stijl misschien de goedkeuring wegdragen van een handjevol fijnproevers die genieten van een droog feitenrelaas, maar slaat ze een modderfiguur ten overstaan van een groot publiek, dat net zo goed vermaakt en gemanipuleerd moet worden. Op deze manier wordt de perfecte redenaar door Cicero gedefinieerd als een meester van de drie stijlen. Dionysius verankert de drie registers op zijn beurt in een complex systeem van stilistische deugden: de simpele stijl bevat alleen ‘essentiële deugden’ (ἀρεταὶ ἀναγκαῖαι ), zoals helderheid en correctheid, terwijl de verheven stijl zich hoofdzakelijk bedient van ‘aanvullende deugden’ (ἀρεταὶ ἐπίθετοι ), zoals emotie en schoonheid. De middenstijl, die bij Cicero de minste aandacht krijgt, is de spil in Dionysius’ driestijlenleer: dit tussenliggende register beoogt volgens Dionysius een uitgebalanceerde combinatie van de essentiële én de aanvullende deugden te zijn. Om uit te leggen waarom juist deze middenstijl de beste is, roept Dionysius meerdere, schijnbaar tegenstrijdige beelden op: nu eens appelleert hij aan een Aristotelisch midden, dan weer stelt hij de middenstijl voor als een homogeen mengsel of juist als een soort heterogene substantie. Het lijkt erop dat Dionysius zich niet bekommert om de consistentie van zijn betoog, zolang zijn argumenten maar steeds bijdragen aan het bereiken van zijn doel, namelijk het optuigen van een stijl die naar het model van Demosthenes een universele zeggingskracht heeft. Dionysius schrijft niet alleen over Griekse literatuur voor een Grieks publiek, maar hij is ook werkzaam in Rome en draagt enkele traktaten op aan Romeinse aristocraten: het is dus niet verwonderlijk dat hij een prozastijl voorstaat die de hele wereld moet aanspreken. Zo wordt duidelijk dat Cicero en Dionysius beiden de drie stijlen zodanig interpreteren dat ze bijdragen aan hun eigen stilistische programma. Overigens leggen hun besprekingen ook een opvallend verschil bloot tussen de esthetische ervaring van Grieken en Romeinen. Op een Romeins forum lijkt de simpele stijl vooral effectief te zijn bij de geletterde sectie van het publiek, en de verheven stijl bij de grote massa. Op een Griekse agora is de situatie echter precies omgekeerd: de hoogopgeleiden genieten van een verheven stijl, en het ongeletterde

289

SAMENVATTING gepeupel van een betoog in een simpel jasje. De theorie van de drie stijlen biedt al met al slechts algemene parameters, waarbinnen Griekse en Romeinse geleerden hun besprekingen kunnen aanpassen aan uiteenlopende omstandigheden en idealen.

Hoofdstuk 4 De akoestiek van proza: ruwe woordschikking Volgens menige antieke stijlcriticus hangt de kwaliteit van proza voor een belangrijk deel af van de ‘ordening van de woorden’ ( σύνθεσις τῶν ὀνομάτων , compositio verborum ): Cicero besteedt de helft van zijn Orator aan dat onderwerp en Dionysius heeft er zelfs een monografie over geschreven, getiteld Over woordschikking . Samen met auteurs als Philodemus, Demetrius, Longinus en Quintilianus gaan ze in op de muzikaliteit van woordschikking: zij menen dat zowel poëzie als proza — in beide gevallen werd steeds hardop voorgelezen — melodie en ritme ontlenen aan de volgorde van de woorden. Er wordt onderscheid gemaakt tussen twee categorieën van akoestische effecten: aan de ene kant van het spectrum staat een soepele stijl van woordschikking die aangenaam in het gehoor ligt, en diametraal daartegenover bevindt zich een ruwe stijl die de oren juist tart. In hoofdstuk 4 staat het laatstgenoemde, oorteisterende compositietype centraal: blijkbaar erkennen de geleerden in Rome niet alleen het belang van klankschoonheid, maar ook dat van afstotelijke geluiden. Het hoofdstuk spitst zich wederom toe op de algemene en particuliere elementen in de werken van Cicero, Dionysius en hun tijdgenoten. Er ligt een stevige gemeenschappelijke basis onder de Griekse en Latijnse discussies over woordplaatsing. In de overgeleverde teksten worden eufonie en kakofonie namelijk allerminst beschouwd als kwesties van smaak. Integendeel, de auteurs beweren uit te gaan van een universele natuurwet voor het combineren van klanken: combinaties die zich aan deze wet houden worden gezien als soepel, en combinaties die tegen de wet indruisen worden gezien als ruw. Om de werking van het natuurlijke gebod te illustreren vertellen Cicero ( De oratore 3.195-197, Orator 173) en Dionysius ( Over woordschikking 11.8-10) een verhaal over een artiest die een verkeerde noot aanslaat of onverhoeds zijn ritme doorbreekt: in zo’n geval begint het publiek volgens hen onverbiddelijk te joelen, ook al kan men zelf meestal niet uitleggen wat er nu precies misging. Cicero en Dionysius wijzen erop dat iets soortgelijks ook geldt in de literatuur: wij bezitten allen een aangeboren muzikale intuïtie, die onze oren gevoelig maakt voor overtredingen van de algemene klankwet. Welke woordcombinaties zijn dan zo tergend voor ons gehoor? Ook hierover bestaat grote consensus onder de antieke geleerden: akoestische ruigheid treedt volgens hen onwillekeurig op, als de woorden door hun

290

SAMENVATTING plaatsing moeilijk uit te spreken zijn, bijvoorbeeld ten gevolge van klinkerbotsing (σύγκρουσις , hiatus ), een zich langzaam voortslepende ritmiek, of een zin die onverwachts (te vroeg, te laat of zonder duidelijke markering) eindigt. Dat Griekse en Romeinse auteurs ondanks de overduidelijke fonetische verschillen tussen hun moedertalen behoorlijk eensluidend zijn over de totstandkoming van soepele en ruwe klanken, toont aan hoe hecht hun gezamenlijke literaire gedachtewereld is. Grieken en Romeinen verschillen echter behoorlijk van mening wat betreft de stilistische toepassingen van pijnlijke akoestiek. Griekse critici betogen doorgaans dat de inhoud van een verhaal of betoog moet worden ondersteund door de klank van de woorden: ruwe woordschikking is volgens deze redenering zeer geschikt voor de nabootsing van zaken als spanning en ellende. In Latijnse retorisch-kritische teksten worden de Griekse inzichten echter vooral in verband gebracht met klankschoonheid: dat brengt met zich mee dat Romeinse schrijvers die zich ijverig storten op de soepele plaatsing van hun woorden kunnen worden beschuldigd van Griekse weekheid en haarkloverij. Zij kunnen zich door een ruige woordschikking distantiëren van hun Griekse collega’s: de stroefheid van hun zinnen dient dan als bewijs dat ze zich concentreren op de inhoud en niet op pietluttige details. Door dit globale verschil tussen Grieken en Romeinen is woordschikking bij uitstek een onderwerp waarover beide groepen met elkaar in dialoog treden: Romeinse retorici zetten zich in verschillende gradaties af tegen de Griekse uitgangspunten, en Griekse critici in Rome reageren vervolgens weer op de stellingname van hun Romeinse stadsgenoten. Daarnaast dient te worden opgemerkt dat Grieken en Romeinen in ruwe woordplaatsingen ook de verre echo van de oudheid menen te horen: de stroefheid van de klanken wordt geassocieerd met een rudimentaire esthetica uit een primitieve tijd. Onze bronnen schotelen ons op dit punt een boeiende discussie voor: sommigen, onder wie Dionysius, prijzen ruwheid als symbool van eerbiedwaardige ouderdom, terwijl anderen, onder wie Cicero, haar juist afwijzen als een achterhaald alternatief voor een modernere, meer verfijnde stijl. Zo gelijkgestemd als de literatuurkenners in Rome zijn over de mechanismes van woordschikking, zo meeslepend zijn hun gevechten over de juiste toepassing ervan.

Hoofdstuk 5 De slag om Attica: prozastijl en politiek in Rome Een invloedrijke antieke wijsheid stelt dat iemands manier van spreken een spiegel is van diens karakter: qualis homo, talis oratio . Naar analogie valt te verwachten dat de stilistische opvattingen van retorici en critici nauw aansluiten bij de andere activiteiten die zij tijdens hun

291

SAMENVATTING leven ontplooien. In hoofdstuk 5 worden daarom de ideeën over stijl van Cicero, Dionysius en de redenaar en neoterische dichter Calvus (ca. 80-ca. 50 v. Chr.) in verband gebracht met hun morele en politieke opvattingen. Deze drie auteurs zijn verwikkeld in een heuse slag om Attica: ook al verschillen zij behoorlijk van mening over de definitie van geweldig proza, zij adverteren niettemin allen hun eigen positie als ‘Attisch’ ( Ἀττικός , Atticus ). Dit kunnen zij doen, doordat Attica in Rome wordt geassocieerd met een breed scala aan literaire, morele en politieke waarden, waaruit de geleerden die elementen kunnen selecteren die aansluiten bij hun eigen doeleinden. Zo verwijst Cicero’s ‘Atticisme’ met name naar de vrije democratie en het krachtige patriottisme van de Atheners, legt Dionysius de nadruk op hun vermeende gezonde verstand en zelfbeheersing, en beklemtoont Calvus hun onbezoedelde puurheid en mannelijkheid. Die gearceerde Attische waarden, afkomstig uit een gemeenschappelijk repertoire, zijn niet alleen leidend in hun verhandelingen over stijl, maar ook in hun morele en politieke beschouwingen over het Rome van hun tijd. In zijn retorisch-theoretische werken uit 46 v. Chr. bepleit Cicero een Attisch stijlideaal dat niet los kan worden gezien van zijn verzet tegen de dictatuur van Caesar (48-44 v. Chr.): in zijn discussie over de drie stijlen spreekt hij zijn voorkeur uit voor het verheven register, omdat de overweldigende kracht van dit stijltype volgens hem resteert als enige wapen tegen de ondergang van de vrije Romeinse republiek. Tegenover het machtsvertoon van Caesar moet Cicero’s ideale redenaar zich niet opstellen als een terughoudende Lysias, maar als een gepassioneerde Demosthenes. De republiek maakt echter onherroepelijk plaats voor een autocratie: onder de regering van keizer Augustus (27 v. Chr.-14 n. Chr.) formuleert Dionysius een eigen interpretatie van Attische welsprekendheid, die inspeelt op de nieuwe politieke realiteit. Zijn pleidooi voor een evenwichtige middenstijl grijpt niet alleen terug op de deugdzame gematigdheid waarmee volgens hem de klassieke Atheense polis en de vroege Romeinse samenleving doordesemd was, maar het sluit ook naadloos aan bij de propaganda van de nieuwe heerser: Augustus presenteert zichzelf immers als de hersteller van de oude moraal en bovendien als een zuinig mens en een gematigd spreker. Het lijdt dus geen twijfel dat prozastijl en politiek onlosmakelijk met elkaar verbonden zijn. Dat punt wordt eens te meer onderstreept door de stijlleer van Calvus, de leider van de zelfverklaarde ‘Attische’ redenaars in Rome, wier anti-Ciceroniaanse retorica reeds in hoofdstuk 3 aan bod is gekomen. Calvus was actief gedurende het triumviraat van Caesar, Pompeius en Crassus (59-53 v. Chr.). Zowel in zijn redevoeringen als in zijn poëzie propageert hij een stijl die vrij is van overdreven gewichtigdoenerij en zijige aanstellerij. Tegelijkertijd drijft hij de spot met de machtige generaals Caesar en Pompeius vanwege hun

292

SAMENVATTING vermeende vrouwelijke rol tijdens homoseksuele escapades: omdat hun manier van handelen niet voldoet aan Calvus’ eisen van viriliteit, kunnen zij rekenen op zijn vileine kritiek. De strenge Romein gaat zelfs nog een stap verder: onze bronnen suggereren dat hij zichzelf onderwierp aan medische therapieën op basis van lood en koud water om zijn eigen mannelijkheid te bewaren. Op basis van breed voorhanden zijnde uitgangspunten komt Calvus tot een zeer persoonlijke stijltheorie die inderdaad een spiegel lijkt te zijn van zijn diepste ziel.

Hoofdstuk 6 Algemene conclusie De vier detailstudies van deze dissertatie over respectievelijk de stilistische modellen Demosthenes en Hegesias (hoofdstuk 2), de theorie van de drie stijlen (hoofdstuk 3), de akoestiek van ruwe woordschikking (hoofdstuk 4) en Atticisme (hoofdstuk 5) bieden niet alleen een overzicht van de belangrijkste thema’s in de stijltheorieën uit laat-republikeins en Augusteïsch Rome, maar ze verlenen tevens geloofwaardigheid aan de drie centrale claims die dit proefschrift te berde brengt: Cicero, Dionysius en de andere retorici en critici uit dezelfde periode (1) nemen deel aan een gemeenschappelijk vertoog over prozastijl dat (2) ruimte biedt aan diverse, soms zelfs tegenstrijdige opvattingen volgens de uiteenlopende programma’s en ambities van de betrokken Griekse en Romeinse auteurs die (3) in een actieve dialoog voortdurend de kaders van hun gedeelde discours herijken. In hoofdstuk 6 worden de belangrijkste argumenten voor deze stellingen nog eens op een rij gezet. Het wijst ten slotte op een toegevoegde waarde van deze dissertatie voor eenieder met een algemene interesse in meeslepende literatuur en overweldigende welsprekendheid: dit onderzoek verschaft inzicht in de voorkeuren, motivaties en doelstellingen die ten grondslag liggen aan de esthetische trends uit een van de meest geroemde literaire en retorische bloeiperioden aller tijden — het Rome van Cicero en Dionysius van Halicarnassus.

293

Curriculum Vitae

Steven Ooms werd geboren op 4 maart 1989 te Breda. In 2007 behaalde hij zijn gymnasiumdiploma aan de christelijke scholengemeenschap Blaise Pascal te Spijkenisse. Vervolgens studeerde hij rechtsgeleerdheid (propedeuse) en klassieke talen aan de Universiteit Leiden. Voor de laatstgenoemde studie verwierf hij in 2010 zijn BA-graad met een scriptie over de Griekse literair-esthetische term ἐναγώνιος (cum laude ). Hij voltooide zijn ResMA-opleiding in 2013 met een scriptie over poëziekritiek in Plato’s Protagoras (cum laude ). In 2019 behaalde hij zijn eerstegraads lesbevoegdheid aan het ICLON, de leraren- opleiding van de Universiteit Leiden ( cum laude ). Samen met dr. Casper de Jonge publiceerde hij een bewerkte versie van zijn BA-scriptie in het internationale tijdschrift Classical Philology . Tussen 2014 en 2018 was Steven Ooms als promovendus verbonden aan het Leiden University Centre for the Arts in Society (LUCAS). Zijn promotieonderzoek, onder supervisie van prof. dr. Ineke Sluiter en dr. Casper de Jonge, maakt deel uit van het NWO-Vidi-project ‘Greek Criticism and Latin Literature: Classicism and Cultural Interaction in Late-Republican and Early-Imperial Rome’. Gedurende zijn aanstelling hield hij verscheidene lezingen over zijn onderzoek en verzorgde hij voor de opleiding klassieke talen onderwijs op het gebied van zowel Griekse als Latijnse taal- en letterkunde: hij gaf onder andere colleges aan eerstejaars studenten over Plato en Cicero. Daarnaast is Steven Ooms vanaf negentienjarige leeftijd actief geweest als docent op verschillende middelbare scholen en instellingen in Spijkenisse, Hellevoetsluis, Rotterdam, Amsterdam en Den Haag. Sinds augustus 2018 is hij werkzaam op het Erasmiaans Gymnasium te Rotterdam.

294